Microsoft Exchange Load Generator 2013

Microsoft® Exchange Load Generator 2013
Microsoft Corporation
Published: October 2013
This document provides information about working with the Exchange Load Generator tool. This
tool helps administrators perform benchmarking, pre-deployment validation, and stress testing
tasks that introduce various types of workloads into a test (non-production) Exchange messaging
system.
Comments? Send feedback to [email protected]
1
Contents
Microsoft® Exchange Load Generator............................................................................................ 1
Microsoft Exchange Load Generator........................................................................................... 5
System Requirements for Exchange Load Generator.................................................................6
When to Use Exchange Load Generator.....................................................................................7
Installing Exchange Load Generator on a Client Computer.........................................................8
Creating and Running the Exchange Load Generator Test.........................................................9
How to Start Exchange Load Generator.................................................................................11
How to Set Up the Exchange Load Generator Test Topology................................................12
How to Configure the Test Properties.....................................................................................14
Customizing the Exchange Load Generator Test...................................................................16
Specifying Threshold Detection during Simulation.................................................................36
How to Enable Remote Control.............................................................................................. 41
How to Enable Stress Mode................................................................................................... 43
How to Enable RPC over HTTP............................................................................................. 43
How to Initialize the Exchange Load Generator Test..............................................................44
How to Run the Exchange Load Generator Test....................................................................46
How to Stop the Simulation Process......................................................................................47
How to Repeat the Test.......................................................................................................... 47
Running Exchange Load Generator as a Command-Line Program..........................................48
Running Exchange Load Generator Tasks................................................................................49
Using System Monitor and Exchange Load Generator Data.....................................................51
How to Consolidate System Monitor Data..............................................................................52
Operational Recommendations for Using Exchange Load Generator.......................................53
Specific Features and Behaviors of Exchange Load Generator................................................57
Public Folders........................................................................................................................ 57
Mail Flow and Recipients....................................................................................................... 57
Task Rate, Concurrency Settings, and Stress Mode..............................................................58
Dynamic Mail Generator........................................................................................................ 59
Performance Counters Reference.............................................................................................62
Scripted Load Modules in the Exchange Load Generator Framework......................................70
Exchange Load Generator Configuration XML.......................................................................71
Base Script Verbs................................................................................................................... 72
The POP3 Module.................................................................................................................. 76
The SMTP Module................................................................................................................. 78
The IMAP4 Module................................................................................................................ 79
The Outlook Web Access Module.......................................................................................... 84
Frequently Asked Questions about Exchange Load Generator.................................................96
Copyright................................................................................................................................. 105
3
Microsoft Exchange Load Generator
Microsoft® Exchange Load Generator is a pre-deployment validation, and stress testing tool that
introduces various types of workloads into a test (non-production) Exchange messaging system.
Exchange Load Generator lets you simulate the delivery of multiple MAPI client messaging
requests to an Exchange server. To simulate the delivery of these messaging requests, you run
Exchange Load Generator tests on client computers. These tests send multiple messaging
requests to the Exchange server, which causes a mail load.
After the tests are complete, you can use the results to help do the following:
 Verify the overall deployment plan
 Identify bottlenecks on the server
 Validate Exchange settings and server configurations
Exchange Load Generator tests the effect of client load on a server.
There are general tasks such as logoff, logon, and send mail that are common to most protocols.
Additionally, there are feature tasks that are available only in certain protocols.
System Requirements for Exchange Load

Generator
To use Microsoft® Exchange Load Generator, you must meet the minimum and recommended
system requirements outlined in the following table.
Component Minimum Recommended
Processor Any of the following Any of the following

processors: processors:
X64 architecture-based x64 architecture-based
processor processor running at 2.2 GHz
or faster
Operating system Microsoft Windows Server Microsoft Windows Server

2008 R2 2008 R2
Microsoft Windows Server
2012
Windows 7
Windows 8
Memory 512 megabytes (MB) of RAM 2 gigabytes (GB) of RAM
Available hard-disk drive 300 MB 300 MB
Display SVGA (800 x 600) or higher SVGA (800 x 600) or higher

resolution monitor resolution monitor
4
Component Minimum Recommended
.NET Framework Microsoft .NET Framework 4.5 Microsoft .NET Framework 4.5
Redistributable Package Redistributable Package
Running against Exchange Exchange Server 2013 Exchange Server 2013

Server
When to Use Exchange Load Generator

Microsoft® Exchange Load Generator simulates the server workload that is generated by user
interaction with various message client software. It is a useful tool for server administrators or
messaging deployment engineers who are sizing servers and validating a deployment plan.
Specifically, Exchange Load Generator helps you determine whether each of your servers can
handle the load that they are intended to carry. Another use for Exchange Load Generator is to
help validate the overall solution. Because Exchange Load Generator simulates client requests,
the effect of server-side solutions such as archiving, antivirus, or anti-spam products can also be
validated.
Important
 Exchange Load Generator should be used only in test environments that have no
connection to the production environment. This tool should not be used in a production
environment, an environment that is mission critical, or one that contains important
information of any kind anywhere in the network.
 Exchange Load Generator uses many simulated user mailboxes to create the server
workload. Because mailboxes must be part of a domain user’s account, the Exchange
Load Generator tool therefore creates many domain user accounts to support these user
mailboxes. By design, Exchange Load Generator requires that the password associated
with these domain accounts be the same. Because this most likely does not comply with
your organization's security requirements, to lessen any risk this could present, we
recommend that this tool be used only on isolated test networks that do not have
connectivity to your production network.
 Because load simulation works by using system resources, Exchange Load Generator is
unsuitable for use on production networks because it could interfere with production
operations by competing for those resources.
Additionally, Exchange Load Generator does not account for all factors that are required to size
servers completely. The following factors are not simulated by Exchange Load Generator and can
affect server capacity planning:
 Loadgen will not replicate the correct IOPs per user. For IOP testing, we recommend you
use Microsoft Exchange Server Jetstress.
5
Exchange Load Generator also does not provide a complete picture of user experience, and its
results should not be interpreted in that manner.
Installing Exchange Load Generator on a Client

Computer
To simulate many clients sending messaging requests to an Exchange server, you must install
and configure Exchange Load Generator on one or more client computers. This topic provides
the installation procedures for each client computer. The following are some important
considerations for installation.
 You can install Exchange Load Generator on an Exchange Server but the actual simulation
should be run from separate client computers.
 If you do not log on as the enterprise administrator (for example if you log on as the local
administrator), Exchange Load Generator cannot create users in Active Directory. Exchange
Load Generator must have sufficient rights to create users, organizational units, distribution
groups, and query-based distribution groups. Exchange Load Generator must be running
under an account that is a member of the Enterprise Administrators group. To view the group
membership of a user, double-click the user account name in the Active Directory Users and
Computers snap-in, and then click Member Of. If you are logged on as a user who has
correct permissions, you can add a user to a security group. To add a user to a security
group, double-click the group in the Active Directory Users and Computers snap-in, click
Members, and then click Add.
 We recommend that you join the client computer to the domain where the Exchange servers
are located. Exchange Load Generator does not require that the computer experience be
domain-joined. However, the user experience will be enhanced if you follow this step.
Follow these steps to install Exchange Load Generator on the client computer.
To install Exchange Load Generator on the client computer

1. Log on to the client computer by using domain credentials. As mentioned, we recommend
that you join the client computer to the domain in which the Exchange servers are
located.
2. Download and extract the Exchange Load Generator files and installer (Loadgen.msi).
3. Double-click Loadgen.msi. The Microsoft Exchange Load Generator Setup Wizard
opens and steps you through the installation. The installer file creates a new folder on the
system drive named Exchange Load Generator, and copies the Exchange Load
Generator files to that new directory. You can either accept the default settings in the
installation or customize the installation location for your requirements.
6
Creating and Running the Exchange Load
Generator Test
First, you will need to configure the storage groups and mailbox stores on the Exchange server.
Additionally, since MAPI connectivity to Exchange 2013 occurs via RPC/HTTP (Outlook
Anywhere) it is critical to properly configure the CAS servers that LoadGen will connect to, as well
as any network infrastructure such as load balancers that may exist between LoadGen clients
and the CAS servers. Once server configuration is completed you can create and run the
Exchange Load Generator tests. This topic highlights the steps that are required to create and
run an Exchange Load Generator test on a client computer. Also included is a table that indicates
which method is used to do the various steps.
Each of these steps is described in detail in accompanying topics.
1. How to Configure Exchange RPC/HTTP. Proper configuration of RPC/HTTP connections

is critical as MAPI is used to initialize mailboxes, as well as for all Outlook task
simulation.
2. How to Start Exchange Load Generator. Start the simulation on the Start menu
(Programs), from the new graphical user interface, or from a command-line program.
3. How to Set Up the Exchange Load Generator Test Topology . To set up the test topology,
specify the number of users you want to simulate for each mailbox store. Also, create the
distribution groups in Active Directory on the Exchange server. You must follow this step
on the first client only.
4. How to Configure the Test Properties. To configure the test properties, specify the length
of time and the range of simulated users who will run on the client computer.
5. Customizing the Exchange Load Generator Test. Optional. To customize the test,
customize the default behaviors of the test tasks, logon processes, and initialization.
6. How to Enable Remote Control. Optional. If you use more than one client for the
Exchange Load Generator test, enable Remote Control through the graphical user
interface so that you can manage all the clients from one computer. You can enable or
disable load balancing as well during the remote control step.
7. How to Initialize the Exchange Load Generator Test . If you use more than one client for
the Exchange Load Generator test and you have enabled Remote Control, you must do
this step from the client that controls the other clients. The initialization process populates
the users with content that is required for running the tests. Only the Outlook client types
can be used to initialize users.
8. How to Run the Exchange Load Generator Test. You must manually start the test on each
client participating in the test or from the Master Loadgen machine.
9. How to Repeat the Test. Optional. You may want to run the same Exchange Load
Generator test to validate the results of your initial test. You may also want to run a test
using slightly different client numbers and characteristics.
7
Methods used to do Tasks
Steps and sub steps Method(s) used
Starting the simulation Start menu, graphical user interface, or

How to Start Exchange Load Generator command-line program
Setting up the topology, which involves Graphical user interface

specifying the number of users for each
mailbox store
How to Set Up the Exchange Load Generator
Test Topology
Configuring the test properties, which involves Command-line program to create file. Edit the
generating and editing a configuration file configuration file.
Running Exchange Load Generator as a
Command-Line Program
Customizing the test by task Edit the configuration file parameters

Customizing the Exchange Load Generator
Test
Common task: Customizing logon and logoff Edit the configuration file
behavior of users
Test
Common task: Enabling stress mode Edit the configuration file

Test
Common task: Customizing the initialization of Edit the configuration file

the test
Test
Enabling remote control Edit the configuration file. Graphical user

How to Enable Remote Control interface.
Setting up the test topology Graphical user interface

Test Topology
Verify that users were created in Active Active Directory Users and Computers
Directory
Test Topology
8
Steps and sub steps Method(s) used
Initializing the test Edit the configuration file

How to Initialize the Exchange Load Generator
Test
Monitoring the progress of the initialization Run Performance Monitor (Perfmon)

Performance Counters Reference
Initializing the mailboxes Edit the configuration file

Test
Initializing public folders Edit the configuration file

Test
Running the test Command-line program. Graphical user

How to Run the Exchange Load Generator Test interface.
Stopping the simulation test Command-line program. Graphical user

How to Stop the Simulation Process interface
Repeating the test Control Panel > Administrative Tools. Active

How to Repeat the Test Directory Users and Computers.
Deleting database and log files Control Panel > Administrative Tools
How to Repeat the Test
Removing users from Active Directory Active Directory Users and Computers
How to Configure Exchange RPC/HTTP

Prior to using Exchange Load Generator in a test topology, you must ensure that RPC/HTTP
(Outlook Anywhere) connectivity is configured properly. This will typically involve configuring a
namespace for Outlook Anywhere connectivity which is consistent across the Client Access
Server (CAS) role machines, selecting an authentication method for Outlook Anywhere, installing
a certificate that is associated with the namespace, configuring load balancing across the CAS
servers, and configuring DNS or the local “hosts” file to map the configured namespace to the
load balanced IP address.
Note:
When an Outlook Anywhere external namespace is configured in the test topology, Exchange
Load Generator will detect that value and use it in the ClientAccessServer list. If Exchange Load
9
Generator is unable to obtain information from the test topology during creation of the initial
configuration (this can occur if the Exchange Load Generator clients are not domain-joined), the
namespace must be manually added to the ClientAccessServer list.
Note:
The following steps which configure a shared external namespace can be skipped if there is only
one Exchange 2013 CAS server (or multi-role server) in the test topology, as long as the
authentication method has been configured to NTLM on the server (or both the server and
Exchange Load Generator have been configured to use Basic authentication), and the self-signed
SSL certificate on the CAS server is fully trusted on all Exchange Load Generator client
machines.
To configure an external Outlook Anywhere namespace

1. Connect to the Exchange Management Shell on one of the Exchange 2013 servers
in the test topology.
2. Utilize the following PowerShell command to configure an external namespace on all
CAS servers, as well as select the NTLM authentication method (replace
“mail.contoso.com” with the correct external namespace for your test deployment).
Get-OutlookAnywhere |Set-OutlookAnywhere -ExternalHostname mail.contoso.com -

DefaultAuthenticationMethod NTLM -ExternalClientsRequireSsl $true
Note:
If the Basic authentication method is used, you must manually adjust the
RpcHttpAuthenticationType parameter (set to “Basic”) in the Exchange Load
Generator configuration file. We recommend using NTLM to simplify the test
configuration process.
Note:
The new settings will not take effect immediately. We recommend either restarting
your Exchange 2013 servers or waiting at least 30 minutes after making these
changes before performing any validation steps.
To install a certificate on CAS servers

1. Find a SSL certificate provider such as the one you intend to use for your production
deployment, or deploy Active Directory Certificate Services on one of the servers in
the test topology.
2. See the Digital Certificates and SSL section of the TechNet Library for details on how
to request and apply a certificate with Exchange 2013.
3. Once the SSL certificate has been deployed on all CAS servers, ensure that all
10
Exchange Load Generator client machines fully trust the certificate authority (CA)
which issued the certificate. If the certificate was obtained from a public CA it is likely
that the CA will already be trusted. Otherwise, obtain the CA certificate and import it
into the “Trusted Root Certification Authorities” portion of the Local Computer
certificate store.
To configure name resolution on Exchange Load Generator client machines via the
hosts file
Note:
If the configured namespace is resolvable via DNS in the test topology, this step can
be skipped.
1. On each Exchange Load Generator client machine, open the Notepad tool.
2. Using Notepad, open the %systemroot%\system32\drivers\etc\hosts file
(%systemroot% is typically c:\windows).
3. Add an entry to the end of the file containing the target IP address followed by the
namespace, such as:
192.168.1.50 mail.contoso.com
4. Save the file and close Notepad.
To validate that Outlook Anywhere connectivity is working

1. Create a test account in the Exchange 2013 test deployment.
2. Install Microsoft Outlook on one or more of the Exchange Load Generator client
machines.
3. Ensure that Outlook can access the test account by configuring a profile, connecting
to the account, and sending test mail to another account.
How to Start Exchange Load Generator

You can start Exchange Load Generator on the Start menu, from the graphical user interface
(LoadGenWin.exe), or from a command-line program (LoadGenCmd.exe).
To start Exchange Load Generator on the Start menu

4. Click Start, and then click All Programs.
5. Open Microsoft Exchange
6. Click Exchange Load Generator 2013.
11
Note:
This procedure starts the graphical user interface version of Exchange Load
Generator.
To start Exchange Load Generator at a command prompt

1. Click Start, click Run, and then type cmd to open a Command Prompt window.
2. Locate the folder in which you installed the Exchange Load Generator files. By default,
this is the Exchange Load Generator directory on the system drive.
3. At the command line, type LoadGenCmd.exe.
For more information about starting Exchange Load Generator at a command prompt, see
Running Exchange Load Generator as a Command-Line Program.
How to Set Up the Exchange Load Generator Test Topology

This topic explains how to specify the number of users to simulate for each mailbox store that
receives Exchange Load Generator requests. If you use more than one client to run the same
Exchange Load Generator test, you have to perform this step on one client only. You do not have
to create the topology on every client. You must configure the number of users in the mailbox
stores where you want to send requests.
After you specify the topology parameters, you have to create the test topology. This procedure
creates the Exchange Load Generator users, distribution groups (also known as DLs or
distribution lists), and dynamic distribution groups (also known as DDLs or dynamic distribution
lists) in Active Directory.
Important
 You must configure Exchange Server 2013 before you create the topology. This could
involve, for example, creating the required databases, pointing databases to the correct
public folders, or changing distribution group settings.
To set up the topology for an Exchange Load Generator test

1. Log on to the client computer.
2. Start Exchange Load Generator by double-clicking LoadGenWin.exe. For more
information, see How to Start Exchange Load Generator.
3. On the Welcome to the Exchange Load Generator page, click Start a new test.
4. On the Start a new test page, create a new configuration or use an existing configuration
file by clicking the selection you want. Click Continue.
5. On the Specify test settings page, define the global test settings as well as the
Topology permissions and user permissions.
12
You can choose “Skip recipient management and continue with user group
configuration” if your users have already been created. Continue to step 8.
If your users have not been created yet and you wish to do so through Loadgen click
“Continue with recipient management” and continue to step 6.
6. On the User settings page, set up according to your test requirements. For example,
select the user count that corresponds to each mailbox database, and then enter the
number of users that you want to simulate.
7. On the Advanced recipient settings page, configure the options of Distribution List
Settings, Dynamic Distribution List Settings, Contact Settings, and External
Recipient Settings according to your test requirements. Click Continue. The progress
bar for Recipient creation is in progress is displayed.
8. When the recipient creation task is completed, the Specify test user groups page will
be displayed. From here you can configure one or more groups by Name, Client Type,
Action Profile, Mailbox Profile, PreTestLogon or Container. You can edit any of the
fields, and you can add or delete user groups as required. Click Continue.
9. The Remote configurations page lets you configure one or more remote load
generators. This is an optional task. Click Continue.
10. A Configuration summary Review this summary, and then select an action from the
selection in the lower half of the screen. This includes going back and adjusting or re-
doing any items that you choose.
To verify that users are created in Active Directory

1. On the Exchange server, start the Active Directory Users and Computers snap-in by
clicking Start, click Programs, double-click Administrative Tools, and then click Active
Directory Users and Computers.
2. In the console tree, locate the domain you want, and then expand that domain.
3. Expand LoadGen Objects/Users and verify that a subfolder with the Exchange server
name exists. The distribution groups are displayed in LoadGen Objects/Distribution
Lists, the dynamic distribution groups are displayed in LoadGen Objects/Dynamic
Distribution Lists, and the contacts are displayed in LoadGen Objects/Contacts. The
following figure shows how everything appears in Active Directory.
13
How to Configure the Test Properties
The procedure explains how to configure the test properties by generating a configuration file and
then editing it. Tasks include setting up user credentials and specifying parameters such as the
length of time that the test should run.
To configure the test properties

1. Log on to the client computer by using domain credentials. If you have not joined the
client computer to a domain, log on with a local administrator account. We recommend
that you use a domain account to enhance the user experience.
2. Open a Command Prompt window and locate the Exchange Load Generator installation
directory.
3. Run the command LoadGenCmd.exe /g which will generate a generic configuration file
that is named LoadGenConfig.xml in the current directory. You will need to specify
Configuration Arguments in order for the LoadGenConfig.xml to be created.
4. Open the LoadGenConfig.xml file by using Notepad or a similar editor. The section in the
XML Configuration File Internals, describes the meaning and purpose of critical
14
parameters listed in the configuration file.
5. Search for the LdapSearchFilter parameter. This parameter is used to restrict the set of
users who will be included in the test run. By default, it includes all mail-enabled users in
the directory at or below the base domain name that is specified in the LdapBaseDn
parameter, which immediately follows LdapSearchFilter. You will notice that
LdapBaseDn has already been set to the correct value if you are running LoadGenCmd
while logged in with domain credentials. If the LdapBaseDn parameter does not correctly
specify the base location of your test users, you will need to manually adjust this
parameter.
6. Search for the TargetForest parameter. This parameter specifies which forest should be
used to find users and Exchange topology information. The value can be any of the
following: a NetBIOS or fully qualified domain name representing a global catalog server,
domain controller, domain in the forest, or root of the forest. This parameter has already
been set to an acceptable value if you are running LoadGenCmd when logged on with
domain credentials. If the value must be changed for any reason, you can edit it at this
point. An example of the xml script is as follows, with an accompanying discussion of the
parameters.
<TargetForest>E12</TargetForest>
<DirectoryUsername>administrator</DirectoryUsername>
<DirectoryPassword>Manager1</DirectoryPassword>
<DirectoryDomain>E12</DirectoryDomain>
<MasterPassword>Manager1</MasterPassword>
<RunLengthDuration>P0Y0M0DT8H0M0S</RunLengthDuration>
<SimulatedDayLength>P0Y0M0DT8H0M0S</SimulatedDayLength>
 The DirectoryUsername, DirectoryPassword, and DirectoryDomain parameters
specify the credentials that will be used to access Active Directory during the test.
You will minimally need to adjust the DirectoryPassword parameter.
 The MasterPassword and MasterDomain parameters specify the domain and
password that can be used to authenticate with each test account. You will minimally
need to set the MasterPassword parameter equal to the password that you specified
when you created the test accounts.
 The RunLengthDuration parameter specifies how long a test will run before
stopping. By default, this is equal to 1 hour. The parameter is specified by using the
ISO 8601 extended format of PaYbMcDTdHeMfS where a equals the number of
years, b equals the number of months, c equals the number of days, d equals the
number of hours, e equals the number of minutes, and f equals the number of
seconds. A value of P0Y0M0DT1H0M0S means 1 hour, as does PT1H. To change
this to an 8-hour test, you could use P8HS as the value. If you want the test to run
with an infinite duration, specify a duration of 0, for example PT0S.
To define RunLengthDuration at each user group, specify this global
RunLengthDuration as 0.
15
7. Save your configuration file.
Customizing the Exchange Load Generator Test

Microsoft Exchange Load Generator simulates some calls that Microsoft Office Outlook makes to
the server that is running Exchange Server. The most common user tasks are combined into one
Exchange Load Generator task named Read and Process Message. With the Read and Process
Message task, the simulated user reads all the e-mail in the Inbox and replies to, replies to all,
forwards, moves, deletes, or loads attachment to a certain percentage of the e-mail messages.
This topic lists the parameters, by task alphabetically, that can be changed in the configuration
file. These tasks include the common tasks for MAPI and Outlook-specific tasks.
Tables of Simulation Tasks and Their Parameters

Generally, to customize the test, you have to make changes in the XML configuration script file.
The following table of tasks, positioned alphabetically by task name, describes the meaning and
purpose of each parameter listed in the configuration file.
Browse Calendar
BrowseCalendar opens the Calendar folder and performs modify or delete actions on content
within the folder. If BrowseCalendarActionCount is 0, no action will be performed on calendar
items.
Parameter Comment
MinimumActionMessages The minimum number of calendar items to

perform an action on.
MaximumActionMessages The maximum number of calendar items to

perform an action on.
ChangePercentage The percentage of calendar items to be

changed.
DeletePercentage The percentage of calendar items to be

deleted.
Browse Contacts
BrowseContacts opens the Contacts folder and performs modify actions on content within the
folder.
There are no parameters for this task.
16
Browse Public Folder
BrowsePublicFolder browses Public Folder content. This action can include opening 1 or more
(1-n) public folders, reading new content, reading 0 or more (0-n) previously read messages,
creating or deleting sub folders, and performing the same processing actions that the
ReadAndProcessMessages task performs on mailbox content.
Parameter Comment
FolderOpenCount The number of folders that will be opened.
OldPostReadCount The number of old posts to read, in addition to

any new posts, in each folder opened.
CreateSubFolderPercentage The percentage of task invocations in which a

new sub folder will be created.
DeleteSubFolderPercentage The percentage of task invocations in which a

sub folder will be deleted.
MinimumFolderDepth The minimum folder depth for the public folder

hierarchy that must be used when you create
sub folders.
MaximumFolderDepth The maximum folder depth for the public folder

sub folders.
AverageFolderDepth The average folder depth for the public folder

sub folders.
ResolveConflictsPercentage The percentage of conflicts that must be

resolved if there are any.
ApplyViewPercentage The percentage of folder opens in which a

random view will be applied.
ReplyPercentage The percentage of processed messages that

will result in a reply to action.
ReplyToFolderPercentage The percentage of processed messages that

will result in a post reply to folder action.
ForwardPercentage The percentage of processed messages that

will result in a forward action.
DeletePercentage The percentage of processed messages that

will result in a delete action.
MovePercentage The percentage of processed messages that

will result in a move action.
17
Parameter Comment
CopyPercentage The percentage of processed messages that

will result in a copy action.
ChangePostPercentage The percentage of processed messages that

will result in a content change action.
MaximumMessages The maximum number of messages that can

remain in any processed folders.
Browse Tasks
BrowseTasks opens the Tasks folder and performs modify or delete actions on content within the
folder. The status of the task must be complete before a deletion is performed.
Create Contact
CreateContact adds a contact to the Contacts folder in the user's mailbox.
Create Folder
CreateFolder creates a folder in the mailbox folder hierarchy. The CreateFolder task can be
tuned to create at the top level of the hierarchy at a given percentage, and also takes minimum,
maximum, and average parameters for the desired hierarchy depth.
Parameter Comment
CreateAtTopLevelPercentage The percentage of folders to be created at the

top level.
MaxDepth The maximum depth of the folder to be created

if it is not at the top level.
AvgDepth The average depth of the folder to be created if

it is not at the top level.
MinDepth The minimum depth of the folder to be created

if it is not at the top level.
Create Tasks
CreateTask adds a task to the Tasks folder in the user's mailbox.
No parameters for this task.
18
Delete Mail
DeleteMail deletes mail from a selected folder, which is controlled by
DeleteFromDeletedItemsPercentage. If you are not deleting from the DeletedItems folder,
Exchange Load Generator randomly selects one from any other folder. The source folder can be
weighted to select the Deleted Items folder.
Parameter Comment
MinimumMessagesToDelete The minimum number of e-mail messages to

be deleted.
MaximumMessagesToDelete The maximum number of e-mail messages to

be deleted.
AverageMessagesToDelete The average number of e-mail messages to be

deleted.
DeleteFromDeletedItemsPercentage The percentage of e-mail messages deleted

from the Deleted Items folder.
Download Offline Address Book

DownloadOAB downloads the offline address book. This operation should simulate downloading
the content based on the usual operation of the simulated client type and should ignore the
received data (that is, do not store it locally).
Parameter Comment
FullDownload Specify whether to always download the Offline

Address Book instead of to make incremental
changes since the last download.
FullDetails Collect full details instead of partial details.
Edit Rules
EditRules modifies the rules in a user's mailbox. This can include create, change, and delete
operations.
Parameter Comment
MinimumRules The minimum number of rules that should exist

in a mailbox.
MaximumRules The maximum number of rules that should exist

in a mailbox.
DeletePercentage The percentage of task invocations in which a
19
Parameter Comment
rule will be deleted.
CreatePercentage The percentage of task invocations in which a

rule will be created.
ChangePercentage The percentage of task invocations in which a

rule will be changed.
ConditionFromDistributionListPercentage The percentage of rules that will be triggered

by a From-Distribution-List condition.
ConditionFromPersonPercentage The percentage of rules that will be triggered

by a From-Person condition.
ConditionSentOnlytoMePercentage The percentage of rules that will be triggered

by a Sent-Only-To-Me condition.
ConditionWithWordInSubjectBodyPercentage The percentage of rules that will be triggered

by a With-Word-In-Subject-Or-Body condition.
ConditionNameInToCcBoxPercentage The percentage of rules that will be triggered

by a Name-In-To-or-Cc-Box condition.
ConditionWithAttachmentPercentage The percentage of rules that will be triggered

by a With-Attachment condition.
ConditionNameNotInToBoxPercentage The percentage of rules that will be triggered

by a Name-Not-In-To-Box condition.
ConditionWithWordInSubjectPercentage The percentage of rules that will be triggered

by a With-Word-In-Subject condition.
ActionDeletePercentage The percentage of rules that result in a delete

action.
ActionMoveToFolderPercentage The percentage of rules that result in a move to

folder action.
ActionCopyToFolderPercentage The percentage of rules that result in a copy to

a folder action.
ActionForwardtoPersonPercentage The percentage of rules that result in a forward

to person action.
ActionForwardtoDistributionListPercentage The percentage of rules that result in a forward

to distribution list action.
ActionPermanentlyDeletePercentage The percentage of rules that result in a

permanently delete action.
ActionStopProcessingMoreRulesPercentage The percentage of rules that result in a stop
20
Parameter Comment
processing more rules action.
Edit Smart Folders

EditSmartFolders modifies the smart folders in a user's mailbox. This can include create,
change, and delete operations.
Parameter Comment
MinimumSmartFolders The minimum number of

smart folders that should
exist in each mailbox.
MaximumSmartFolders The maximum number of

smart folders that should
exist in each mailbox.
DeletePercentage The percentage of task

invocations that should
result in a delete action.
CreatePercentage The percentage of task

result in a create action.
BrowsePercentage The percentage of task

result in a browse action.
Be aware that the
parameters specified for
the browse folder task will
be used if a browser
action is selected.
CriteriaUnreadPercentage The percentage of smart

folders that should use
Unread criteria.
CriteriaImportantPercentage The percentage of smart

Important criteria.
CriteriaOldPercentage The percentage of smart

Old criteria.
CriteriaForFollowUpPercentage The percentage of smart
21
Parameter Comment

For-Follow-Up criteria.
CriteriaConversationWithRandomPersonOrDistributionListPercentage The percentage of smart

Conversation-With-
Person-Or-Distribution-
List criteria.
CriteriaUnreadOrForFollowUpPercentage The percentage of smart

Unread-or-For-Follow-Up
criteria.
CriteriaReceivedThisWeekPercentage The percentage of smart

Received-This-Week
criteria.
CriteriaFromRandomPersonOrDistributionListPercentage The percentage of smart

From-Person-or-
Distribution-List criteria.
CriteriaSentToRandomDistributionListPercentage The percentage of smart

Sent-to-Distribution-List
criteria.
CriteriaWithSpecificWordsPercentage The percentage of smart

With-Specific-Words
criteria.
CriteriaLargePercentage The percentage of smart

the Large criteria.
Export Mail
ExportMail simulates the effect of exporting the content of a folder from the mailbox. The
ExportMail task can be configured to include subfolders of the selected folder, or the whole
mailbox.
22
Note:
This action does not actually result in content being stored locally on the client as that will
not scale. The task only simulates the effect of the server of performing this action.
Parameter Comment
ExportEntireMailboxPercentage The percentage of the whole tree exported.

Otherwise randomly select a top level folder
and export.
IncludeSubFolder Specify whether to include subfolders in export.
Logoff
Logoff disconnects from the server in such a way that client and server side resources
associated with the user connection are released.
Parameter Comment
EmptyDeletedItems Specify whether the Deleted Items Folder will

be emptied on logoff.
Logon
Logon establishes a user connection to the server. Depending on the type of client being
simulated, this can involve more than one connection being established, and can involve any
processing that is typically run as part of starting the client or performing a user initiated action
within the client to establish a connection to the server. For example: authentication, fetching any
information that is required to build the user interface, and so on.
Parameter Comment
AutoDiscoveryPercentage The percentage of autodiscovery performed

during logon.
Make Appointment
MakeAppointment adds an appointment to the calendar folder in the user's mailbox.
Parameter Comment
MinimumAppointmentHours The minimum length of appointment in hours.
MaximumAppointmentHours The maximum length of appointment in hours.
AverageAppointmentHours The average length of appointment in hours.
23
Parameter Comment
RecurringPercentage The percentage of meetings that will have a

random recurrence pattern.
AllDayPercentage The percentage of meetings that will be marked

as all day.
OOFAppointmentPercentage The percentage of appointments that will be

marked as OOF.
Move Mail
MoveMail moves messages from a selected folder to another selected folder in the mailbox. The
source folder selection can be weighted to select the Inbox, and the destination folder can be
weighted to select the Deleted Items folder.
Parameter Comment
MinimumMessagesToMove The minimum number of e-mail messages to

be moved.
MaximumMessagesToMove The maximum number of e-mail messages to

be moved.
AverageMessagesToMove The average number of e-mail messages to be

moved.
MoveToDeletedItemsPercentage The percentage of e-mail messages moved to

Deleted Items
MoveFromInboxPercentage Percentage of e-mail messages moved from

the Inbox folder.
Post Free Busy

PostFreeBusy simulates the action of publishing free/busy data for clients that perform this
behavior.
Note:
This task will not be run if the user's mailbox is running Exchange 2013.
Parameter Comments
UpdateFreeBusyInformation Specify whether to update public Free Busy

information.
MonthCount Specify how many calendar months of Free

Busy information to publish.
24
Public Folder Post
PublicFolderPost post a single item to a Public Folder.
Parameter Comments
MessageFiles Specify that custom message files be used for

this task.
Read and Process Messages

ReadAndProcessMessages simulates the action of a user processing mail in the mailbox.
Processing mail typically consists of reading and responding to unread items. This task will select
a folder to process with a given weight for the Inbox folder. The task will optionally apply a view to
the folder that has been selected. All unread content (or a maximum number of unread
messages) will be read. Some of this content may be left open if a particular client and protocol
support that. For each read message, various actions will be applied based on configured
percentages such as reply, reply all, forward, move, delete, permanent delete, and copy.
Additionally, Incoming meeting requests will be processed by this task.
Parameter Comment
ReplyPercentage The percentage of messages processed that

will result in a Reply action.
ReplyAllPercentage The percentage of messages processed that

will result in a Reply All action.
Forward Percentage The percentage of messages processed that

will result in a Forward action.
MovePercentage The percentage of messages processed that

will result in a Move action.
DeletePercentage The percentage of messages processed that

will result in a Delete action.
PermanentDeletePercentage The percentage of messages processed that

will result in a Permanent Delete action.
CopyPercentage The percentage of messages processed that

will result in a Copy action.
ApplyViewPercentage The percentage of task invocations in which a

random view on the folder will be attached.
LoadAttachementsPercentage The percentage of messages opened that

contain attachments in which the attachments
will be opened.
25
Parameter Comment
PreviewItems Specify whether to preview items in the folder

without opening them.
RespondToMeetingsPercentage The percentage of responses to a meeting

request.
AcceptMeetingPercentage The percentage of meeting requests that will

result in an acceptance.
MaximumOpenMessages The maximum number of messages that can

be kept open.
KeepMessagesOpenPercentage The percentage of messages that can be kept

open.
MaximumFolderMessages The maximum number of messages that can

remain in the folder when the task has
completed. Note that Exchange Load
Generator operates only on a number that is
great than zero.
ProcessInboxPercentage The percentage of processing an Inbox.
MaximumMessagesToProcess The maximum number of unread messages

that will be processed during the task.
Request Meeting
RequestMeeting sends a meeting request message to a set of recipients.
Parameter Comment
MinimumMeetingHours The minimum meeting length in hours.
MaximumMeetingHours The maximum meeting length in hours.
AverageMeetingHours Average meeting length in hours.
MinimumMeetingAttendees The minimum number of attendees to invite to

meetings.
MaximumMeetingAttendees The maximum number of attendees to invite to

meetings.
AverageMeetingAttendees The average number of attendees to invite to

meetings.
AddDistributionListPercentage The percentage of task invocations in which a

distribution list will be added to the attendee list
26
Parameter Comment
of the meeting request.
AddDynamicDistributionListPercentage The percentage of task invocations in which a

dynamic distribution list will be added to the
attendee list of the meeting request.
AddExternalRecipientPercentage The percentage of task invocations in which an

external recipient will be added to the attendee
list of the meeting request.
ExternalContactPercentage The percentage of contacts picked as external

recipients.
AvailabilityCheckPercentage The percentage of attendee availability checks.
Search
Search performs a content search on a folder or throughout the whole mailbox. This results in the
default search behavior for the client platform. For example, on a cached mode Outlook client,
this would be a no-op. For an online mode Outlook client, this would call into the content index
(CI) on an Exchange 2013 server.
Synchronize Folders
SynchronizeFolders simulates the effect of the server during a synchronize folders operation.
This should simulate the actual effect on the server that would be caused by synchronizing a local
version of mailbox data with the data that is stored on the server that uses whatever method is
appropriate for the given client type and protocol.
MailboxStoreProfile and PublicStoreProfile and Their Parameters

The MailboxStoreProfile and the PublicStoreProfile define the mailbox initialization and public
store initialization respectively. The procedures for working with them are in the topic How to
Customize the Initialization of a Test.
MailboxStoreProfile
Parameter Comment
InboxUnreadPercentage The percentage of messages that will be

marked as unread in the Inbox folder.
NewFolderUnreadPercentage The percentage of messages that will be

marked as unread in the new folders.
BuildInboxRules Specify whether to create rules. If set to false,

Exchange Load Generator does nothing.
27
Parameter Comment
Additionally, if there are existing rules in the

mailbox, they are kept there. If this option is set
to true, the FixUpInboxRules option is ignored.
FixUpInboxRules Specify whether to fix up Inbox rules. If set to

true, Exchange Load Generator follows the
existing rules and tries to fix the condition and
actions by setting the valid recipients and
folders. This option takes effect only if
BuildInboxRules is set to false. Generally this
option is used when the mailboxes are restored
by using production data.
InboxMessageCount The number of e-mail messages that will be

created in the Inbox folder.
DeletedMessageCount The number of e-mail messages that will be

created in the Deleted Items folder.
NewFolderCount The number of new folders that will be created

in a mailbox.
NewFolderMessageCount The number of e-mail messages that will be

created by the Exchange Load Generator.
SmartFolderCount The number of smart folders that will be

created.
InboxRuleCount The number of rules that will be created in each

mailbox.
AppointmentCount The number of appointments that will be

created in each calendar.
ContactCount The number of contacts that will be generated

in the Contacts folder.
UseOrdinalFolderName Specify whether to use the ordinal number for

folder name. If set to true, the folder name will
resemble Folder1, Folder2, and so on.
Otherwise, the GUID will be used for folder
names.
InitializeFreeBusy Specify whether to generate a free/busy status

message on an appropriate public folder store
for each mailbox
InitializeIndices Specify whether to pre-create the standard
28
Parameter Comment
indexes, for example, for Contacts.
InitializeCachedModeIndices Specify whether to pre-create the cached mode

indexes.
MessagePercentageRecent The percentage of messages that have a

delivery date that is within the time frame
defined by the
<MessageRecentContentTimeFrame> tag.
MessageRecentContentTimeFrame The period of time frame percentage that

messages within a delivery time, within this
period, are classified as recent.
MessageContentTimeFrame The period of time frame percentage that the

delivery times of all messages are within this
time frame.
AppointmentPercentageRecent The percentage of calendar items that are

scheduled within the time frame that is defined
by the
<AppointmentRecentContentTimeFrame> tag.
AppointmentRecentContentTimeFrame A time frame from the last day of the current

month backward. All calendar items scheduled
within this period are classified as recent. The
default value is two months.
AppointmentContentTimeFrame The period of time from the last day of the

current month that all calendar items are
scheduled within this time frame. The default
value is two months.
TaskCount The number of tasks in the users' mailboxes.

The task completion percentage is defined by
<BrowseTasksActionParameters> under
<ActionProfile>.
PublicStoreProfile
Parameter Comment
TopLevelFolderCount The number of folders at the top level of the

hierarchy.
NonTopLevelFolderCount The number of folders that are not at the top
29
Parameter Comment
level of the hierarchy.
MinimumFolderDepth The minimum depth of folder hierarchy in the

public folders.
AverageFolderDepth The average depth of folder hierarchy in the

public folders.
MaximumFolderDepth The maximum depth of folder hierarchy in the

public folders.
MinimumMessageCount The minimum number of e-mail messages in

the public folders.
AverageMessageCount The average number of e-mail messages in the

public folders.
MaximumMessageCount The maximum number of e-mail messages in

the public folders.
UseOrdinalFolderName Specify whether to use the ordinal number for

folder name. If set to true, the folder name will
resemble Folder1, Folder2, and so on.
Otherwise the GUID will be used for folder
names.
How to Customize the Logon/Logoff Behavior

This optional procedure explains how to customize the logon/logoff behavior of Exchange Load
Generator users.
To customize the logon/logoff behavior of Exchange Load Generator users

1. Open the LoadGenConfig.xml file by using Notepad or a similar editor.
2. To log on immediately at the start of the test, provide the <PreTestLogon> tag with a
value. For example: <PreTestLogon>true</PreTestLogon>.
3. To empty the Deleted Items folder when you log off, use the applicable LogOff
parameter.
How to Customize Simulation Duration for Individual User Groups

These optional procedures explain how to define the wait duration and the simulation duration for
an individual user group.
30
To define the wait duration for a user group
2. Navigate to the user group of your choice.
3. Set the <StartWaitDuration> tag to the duration of your choice. The user group will not
run simulation until the defined duration has passed after the engine has started.
To define the simulation duration for a user group

2. Navigate to the user group of your choice.
3. Set the <RunLengthDuration> tag to the duration of your choice. The user group will run
simulation for the defined duration.
Note:
We recommend that you set global <RunLengthDuration/> to 0 if you are going to
customize durations for individual user groups. If global <RunLengthDuration/> is
defined, individual user groups cannot have a longer duration than the global
duration. Otherwise, the duration for the individual user group will be overwritten by
the global duration.
How to Customize the Initialization of a Test

This optional procedure outlines how to customize the initialization of a test that involves both the
user mailbox and public store initialization.
For more information about the parameters for these tags, see Customizing the Exchange Load
Generator Test.
To customize the initialization of the test

2. For user mailbox initialization, customize the <MailboxStoreProfile> section in the
configuration file, as follows:
<MailboxStoreProfile>
<SortField>2</SortField>
<TaskConfigList>
<TaskConfig d6p1:type="InitializeMailboxTask"
xmlns:d6p1="http://www.w3.org/2001/XMLSchema-instance">
<TaskCount>1</TaskCount>
<TaskParameters d6p1:type="InitializeMailboxTaskParameters">
<InboxUnreadPercentage>0</InboxUnreadPercentage>
<NewFolderUnreadPercentage>0</NewFolderUnreadPercentage>
31
<BuildInboxRules>true</BuildInboxRules>
<FixUpInboxRules>false</FixUpInboxRules>
<InboxMessageCount>42</InboxMessageCount>
<DeletedMessageCount>6</DeletedMessageCount>
<NewFolderCount>5</NewFolderCount>
<DeepFolderCount>0</DeepFolderCount>
<DeepFolderDepth>0</DeepFolderDepth>
<NewFolderMessageCount>4</NewFolderMessageCount>
<ConversationCount>0</ConversationCount>
<MinConversationSize>2</MinConversationSize>
<MaxConversationSize>2</MaxConversationSize>
<AvgConversationSize>2</AvgConversationSize>
<MinConversationFolderSpan>1</MinConversationFolderSpan>
<MaxConversationFolderSpan>1</MaxConversationFolderSpan>
<AvgConversationFolderSpan>1</AvgConversationFolderSpan>
<SmartFolderCount>3</SmartFolderCount>
<KeepExistingSmartFolders>false</KeepExistingSmartFolders>
<InboxRuleCount>8</InboxRuleCount>
<AppointmentCount>25</AppointmentCount>
<ContactCount>30</ContactCount>
<UseOrdinalFolderName>false</UseOrdinalFolderName>
<InitializeFreeBusy>true</InitializeFreeBusy>
<InitializeIndices>true</InitializeIndices>
<InitializeCachedModeIndices>false</InitializeCachedModeIndices>
<MessagePercentageRecent>0</MessagePercentageRecent>
<MessageRecentContentTimeFrame>-P0Y0M0DT0H0M0S</MessageRecentContentTimeFrame>
<MessageContentTimeFrame>P0Y2M0DT0H0M0S</MessageContentTimeFrame>
<AppointmentPercentageRecent>0</AppointmentPercentageRecent>
<AppointmentRecentContentTimeFrame>-P0Y0M0DT0H0M0S</AppointmentRecentContentTimeFr
ame>
<AppointmentContentTimeFrame>P1Y0M0DT0H0M0S</AppointmentContentTimeFrame>
<NewAppointmentWithReminderSetPercentage>0</NewAppointmentWithReminderSetPercentag
e>
<FixUpCalendarReminders>false</FixUpCalendarReminders>
32
<MessageFiles />
</TaskParameters>
</TaskConfig>
</TaskConfigList>
<WellKnownProfileType>2 MB</WellKnownProfileType>
</MailboxStoreProfile>
3. For public store initialization, customize the <PublicStoreProfile> section in the

configuration file as follows:
<PublicStoreProfile>
<TaskConfigList>
<TaskConfig d4p1:type="InitializePublicStoreTask"
<TaskParameters d4p1:type="InitializePublicStoreParameters">
<MessageFiles/>
<TopLevelFolderCount>1</TopLevelFolderCount>
<NonTopLevelFolderCount>1</NonTopLevelFolderCount>
<MinimumFolderDepth>1</MinimumFolderDepth>
<AverageFolderDepth>2</AverageFolderDepth>
<MaximumFolderDepth>3</MaximumFolderDepth>
<MinimumMessageCount>5</MinimumMessageCount>
<AverageMessageCount>10</AverageMessageCount>
<MaximumMessageCount>15</MaximumMessageCount>
<UseOrdinalFolderName>false</UseOrdinalFolderName>
</TaskParameters>
</TaskConfig>
</TaskConfigList>
<WellKnownProfileType>Default</WellKnownProfileType>
</PublicStoreProfile>
How to Customize the Concurrency Profile

This optional procedure outlines how to customize the concurrency profile to define the
participation percentage during different periods throughout the day.
33
To customize the concurrency profile
2. Customize the concurrency profile by adding, removing, or modifying the
<ConcurrencyListEntry> tag in the configuration file. For example, the following
concurrency profile defines two <ConcurrencyListEntry> tags with different concurrency
percentages for different durations in the day.
<ConcurrencyProfile>
<TaskConfigList>
<TaskConfig d6p1:type="ConcurrencyTask"
<TaskParameters d6p1:type="ConcurrencyParameters">
<ConcurrencyTimeline>
<ConcurrencyListEntry>
<Duration>P0Y0M0DT2H0M0S</Duration>
<ConcurrencyPercent>100</ConcurrencyPercent>
<Label>First 2 hours at 100% concurrency</Label>
<UserSessionMinDuration>P0Y0M0DT4H0M0S</UserSessionMinDuration>
<UserSessionAvgDuration>P0Y0M0DT4H0M0S</UserSessionAvgDuration>
<UserSessionMaxDuration>P0Y0M0DT4H0M0S</UserSessionMaxDuration>
</ConcurrencyListEntry>
<ConcurrencyListEntry>
<Duration>P0Y0M0DT8H0M0S</Duration>
<ConcurrencyPercent>80</ConcurrencyPercent>
<Label>Remaining 6 hours at 80% concurrencyLabel>
<UserSessionMinDuration>P0Y0M0DT4H0M0S</UserSessionMinDuration>
<UserSessionAvgDuration>P0Y0M0DT4H0M0S</UserSessionAvgDuration>
<UserSessionMaxDuration>P0Y0M0DT4H0M0S</UserSessionMaxDuration>
</ConcurrencyListEntry>
</ConcurrencyTimeline>
</TaskParameters>
</TaskConfig>
</TaskConfigList>
34
</ConcurrencyProfile>
How to Enable Remote Control

To run Exchange Load Generator from multiple clients, you can use the remote control feature to
manage all the clients from one computer (called a master computer). You do this by changing
the configuration file or by using the graphical user interface.
Note:
Make sure that you start the Microsoft Exchange Load Generator Remote Agent service
on the remote load generator before you start the simulation.
To enable remote control on two client computers in the configuration file

1. In the LoadGenConfig.xml file (or whatever you named the file), locate the
<RemoteProfile> tag followed by the <EnableRemoteDistribution> tag.
2. Make sure that the value between the start tags and the end tags is "true".
3. Change the file as follows (this sample uses RemoteMachine1 and RemoteMachine2 as
the LoadGen remote machines):
<RemoteProfile>
<EnableRemoteDistribution>false</EnableRemoteDistribution>
<EnablePerUserGroupDistribution>false</EnablePerUserGroupDistribution>
<RemoteLoadGenerators>
<RemoteLoadGenerator Name="RemoteMachine1">
<UserGroups />
</RemoteLoadGenerator>
<RemoteLoadGenerator Name="RemoteMachine2">
<UserGroups />
</RemoteLoadGenerator>
</RemoteLoadGenerators>
<LogLevel>Warning</LogLevel>
<PerfmonCountersRefreshTimeinSeconds>5</PerfmonCountersRefreshTimeinSeconds>
</RemoteProfile>
Note:
Although the master computer can also be configured as a load generator, we do
not recommend that you do this. This is because the Exchange Load Generator
counters will no longer be accurate.
35
To enable remote control in the graphical user interface
1. Open the application.
2. On the Remote configuration page, click Enable task engine to distribute the load to
remote load generators.
3. On the Remote configuration page, click Enable load balancing on remote load
generators to enable or disable load balancing.
4. From the drop-down list in Remote logging level, select the log level that you want.
5. Add or delete load generators using the Remote Load Generators selection box.
To enable logging on the remote load generator

1. In the LoadGenConfig.xml file (or whatever you named the file), locate the
<RemoteProfile> tag, and then the <LogLevel> tag that follows it.
2. The default value of LogLevel is Off. This means no log file will be generated on the
remote load generator. To enable the log file, type any of the following values depending
on the log levels that you want: Information, Error, Warning, Verbose.
 Information refers to informational messages, warnings, and error-handling
messages.
 Error refers to output error-handling messages.
 Warning refers to output warnings and error-handling messages.
 Verbose refers to output all debugging and tracing messages.
How to Enable Stress Mode

This optional procedure explains how to enable stress mode.
To enable stress mode

2. Enable stress mode by setting the <StressMode> tag to true.
How to Initialize the Exchange Load Generator Test

Before you run a Microsoft® Exchange Server Load Generator test, you must initialize the
mailboxes that will be used for the test. The initialization process stores much data in each
mailbox based on the <MailboxStoreProfile> tag in the configuration file. You should only run
initialization from one computer at a time or use remoting.
However, if search indexing is running on the server, it competes with Exchange Load Generator
for CPU resources. Depending on whether your goal is to run a highly accurate test or not, you
36
can either disable the search service if you are not concerned about accuracy. Or if you do, you
have two choices for running the Exchange Load Generator initialization.
 Run Exchange Load Generator initialization on the server with search service disabled during
initialization. As soon as the initialization has finished, enable the search service and let it
complete the indexing process all at one time. Be aware that this activity could take a very
long time.
When you initialize the test, Exchange Load Generator adds or deletes messages in the Inbox
and folders of each user in the test so that each mailbox has the number of messages specified
in the <MailboxStoreProfile> tag. Calendar items and Contacts are handled similarly. Therefore
you can re-run initialization to set the mailboxes back to how they were before you ran the test.
Note:
Initialization can take several hours, or even a full day(s), to finish.
The procedure in this topic describes how to initialize the test using the command line and the
configuration file. When you use this method, you must run a separate command for the mailbox
store and for the public store.
If you use the graphical user interface, the mailbox store and public store initialization is
automatically done together. You can step through the user interface until you are at the
Configuration summary page, where you are given two choices:
 Start the initialization phase. We recommend that you do this before you run the test.
 Skip the initialization phase and run the simulation immediately
To initialize the test

1. Log on to the client computer by using domain credentials. If you have not joined the
that you use a domain account because the user experience will be enhanced.
2. Open a command shell and locate the Exchange Load Generator installation directory.
Run the command LoadGenCmd.exe /r /c <configurationfile>, where configurationfile
is the name of your configuration file. By default, this is LoadGenConfig.xml in the current
directory if you have run LoadGenCmd.exe /g. Therefore, if you have not renamed your
configuration file, the command is LoadGenCmd.exe /i.
Important
To view detailed trace information about the actions that the tool is doing, you can add a trace
listener as a command-line option. Trace listeners can be files. Alternatively, they can be the
console if you want to view the output on screen. To view the output on screen, add the
option /add console. To output the trace.log, add the /add trace.log option at the command
line. For example:
C: \Program Files\Exchange Server Load Generator>LoadGenCmd.exe /i
/add trace.log
Microsoft Exchange Load Generator
Configuring LoadGen task engine…
37
Load settings…
NOTE: There are no trace listeners found. You can add/remove shared
trace listeners.
Starting mailbox initialization…
3. Monitor the progress of initialization tasks by using performance counters. Click Start,
click Run, type perfmon, and then click OK.
4. Add the following performance counters to your view:
 Exchange Load Generator(_Total)\Task Queue Length
 Exchange Load Generator(_Total)\Tasks Completed
 Exchange Load Generator(_Total)\Tasks Dispatched
The Tasks Dispatched counter equals the number of users whom you have
configured for the test. One initialization task is dispatched for each user during the
initialization process. The Tasks Completed counter shows the number of initialization
tasks that have been completed. This value slowly increases to equal the value of the
Tasks Dispatched counter until the initialization process is complete. The Tasks
Queue Length counter displays the number of tasks that remain to be run. This value
goes to zero when the initialization process is finished.
5. To cancel the initialization process, see How to Stop the Simulation Process.
Note:
During initialization, Exchange Load Generator will stop if it encounters errors
such as network errors or errors connecting to the Exchange server or
Active Directory. If you restart the initialization after you resolve the errors,
Exchange Load Generator restarts with the first user. It does not delete and re-
create the messages already in the Inbox and folders. This applies also to
Calendar items and Contacts.
6. If you decide to initialize public folders, run the command LoadGenCmd.exe /p /c
<configuration file>, where configuration file is the name of your configuration file.
Exchange Load Generator randomly determines the number of subfolders and messages
in each folder, based on the minimum, maximum, and average parameters that are set in
<PublicStoreProfile>.
When Exchange Load Generator determines the number of subfolders and messages to
create, it will create them if the number of existing folders is less. Alternatively, Exchange
Load Generator will delete some existing public folders if the number of existing folders is
more. If Exchange Load Generator stops the initialization, when you run it again and
decide to initialize public folders again, the number of subfolders and messages
Exchange Load Generator creates will not necessarily match the numbers from the
previous initialization. We recommend that you delete all existing public folders before
you run initialization again.
38
How to Run the Exchange Load Generator Test
A typical test that has an average number of users is set to take ten hours as the duration of the
simulation. Allow for at least two hours for a sufficient warm-up. After this warm-up period, the
clients and the server operate at a sustained baseline because the intensive startup processes
have completed. After the server starts and code is paged into memory, the server operates in a
consistent pattern. If the mailbox store database is new, more database activity can occur as
index pages split. The server does not reach a sustained baseline for adequate benchmarking
until this activity level decreases.
You should collect test information when the server is in this consistent state. After the server has
reached a sustained baseline, it must run long enough to collect statistically significant data. The
time this takes depends on the complexity of the user model and the cumulative rate at which
user actions are being generated. Perform a long calibration test run before you run your actual
tests. For example, if you perform an eight-hour test run and find variation from hour to hour, but
not between two-hour segments, the results are consistent. This indicates that your test runs
should include at least two hours where the server has reached a sustained baseline of
consistent behavior.
For information about performance counters and running Performance Monitor (Perfmon) so that
you can monitor the progress of the simulation, see Performance Counters Reference.
To run the Exchange Load Generator test

1. Start Exchange Load Generator. For the three methods to start the test, see How to Start
Exchange Load Generator.
2. Log on to the client computer that has domain credentials. If you have not joined the
that you use a domain account because the user experience will be enhanced.
3. Open a command shell and locate the Exchange Load Generator installation directory.
4. Run the command LoadGenCmd.exe /r /c <configurationfile>, where configurationfile
is the name of your configuration file. By default, this is LoadGenConfig.xml in the current
directory. If you have not renamed your configuration file, you only need to run
LoadGenCmd.exe /r.
Important:
To view detailed trace information about the actions that the tool is performing,
you can add a "trace listener" as a command-line option. Trace listeners can be
either files or the console, if you want to view the output on screen. To view the
output on screen, add this option to the command line: /add console. To output
to a file that is named trace log, add this option to the command line: /add
trace.log.
5. For running a test using the graphical user interface, after you have opened the tool,
follow the instructions on the Welcome page to Start a new test.
39
How to Stop the Simulation Process
From the graphical user interface, on the Simulation in Progress screen, click Stop Simulation.
From the command line, to cancel the process from the keyboard, press the keyboard shortcut
Ctrl+C. The task engine will be stopped as soon as possible. To force the task engine to stop
immediately, press Ctrl+C repeatedly until you are returned to a command prompt.

This is an optional step. It is common to run subsequent tests using different client numbers and
characteristics. Running subsequent tests gives you a better idea of the capabilities of the
Microsoft Exchange server. Be prepared to rerun your test to validate the test results. Also, be
sure to allocate enough time in your test cycles to do this.
Running Exchange Load Generator as a

Command-Line Program
There are several tasks or actions that you can run for Exchange Load Generator by using the
command-line program LoadGenCmd.exe. One of the key tasks is to generate the generic
configuration file. You must do this in order to set up the adjustable parameters for the test.
When you start Exchange Load Generator at a command prompt, you must indicate if you want
to initialize users or run a simulation. You cannot do these tasks from the same command. You
must run Exchange Load Generator separately for each task. Additionally, you must create user
accounts before generating a configuration file.
To run Exchange Load Generator at a command prompt, locate the LoadGenCmd.exe program,
which is in the directory where Exchange Load Generator was installed. Click Start, click Run,
and type cmd. If it is required, change the directory path to the path of the LoadGenCmd.exe
program, and continue with running the applicable tasks.
The syntax for running Exchange Load Generator as a command-line program is as follows:
LoadGenCmd [Action Argument] [Configuration Argument]
The configuration argument is Config [filename] = Config file (default is /c) LoadGenConfig.xml.
The following table presents action arguments with their examples of use.
Action Argument Example of use Description
Help /h or /? The help for the command-line

program
LoadCtr /l counter.xml Load performance counters
UnloadCtr /u counter.xml Unload performance counters
AddTrace /add trace log Add trace listeners
SupTrace /sup console Suppress trace listeners
40
Action Argument Example of use Description
EnsMessage /ens *.msg Make sure that message files

are formatted correctly and
can be opened successfully
GenConfig /g Generate a generic

configuration file
InitPrivStore /i Initialize mailbox stores
PrepPublicStore /p Prepare the public store
RunLoadSim /r Run load simulation
Running Exchange Load Generator Tasks

The default task parameters for Microsoft® Exchange Load Generator are set up in the
LoadGenConfig.xml file. This can be generated by using the command LoadGenCmd.exe /g.
To customize the test tasks, you have to edit the configuration file and save it to a directory of
your choice. You can also rename it if you choose.
Access the configuration file in either of the following two ways.
 From the command-line tool: LoadGenCmd.exe /r /c <path>\<config.xml>
 From the graphical user interface: After you have started the tool by clicking
LoadGenWin.exe in the installation directory, select the file from the Start a new test
screen.
This topic includes information about how to set up the task information from the command line
(LoadGenCmd.exe) and from the graphical user interface (LoadGenWin.exe).
Setting Parameters using LoadGenCmd and the Configuration

File
After you run LoadGenCmd to generate a new configuration file, you then make any required
changes in this generated file for the test simulation run.
Note:
You can edit the XML file by using an editor of your choice. To download XML Notepad
2007, go to the XML Downloads page at http://go.microsoft.com/fwlink/?linkid=8078.
A typical scenario for setting parameters is as follows:
 Generate a new configuration file by using LoadGenCmd.exe /g.
 Open the configuration file and change parameters as required. Important XML tags include
<ActionProfile> and <ActionCount>.
 There is an <ActionProfile> tag for each defined user group. This tag indicates the start
of the XML that is related to tasks and their parameters, for a particular user group.
41
 In the XML script, a task is of the format <tasknameActionCount>. Therefore, for the
logon task, this would be <LogonActionCount>.
 Additionally, <ActionProfile> has the WellKnownProfileType attribute, which will contain a
value like Outlook_200 or Custom (same as the UI drop down). If you select the Custom
profile, the task parameters can be changed according to the user requirements of the
simulation.
Caution:
We encourage you to use the well-known profiles that are provided by Exchange
Load Generator. Avoid creating a custom profile unless you are very familiar with
Exchange Load Generator parameters, and you understand what type of
difference your customization will make in terms of the load that is generated on
the server.
Running the Simulation Using LoadGenCmd.exe

When the action profile has been set, run LoadGenCmd.exe /r /c config.xml. This runs the
simulation based on the configuration.
Selecting the Action Profile Using LoadGenWin.exe

When you run LoadGenWin.exe, you are provided with the user interface with which to select
the Action Profile type. As you step through the user interface, in the Specify test user groups
page, under Action Profile, select Custom.
However, there is no user interface for editing task parameters so if you want a custom profile, do
the following:
1. On the Configuration summary page, save the configuration file.
2. Close the Microsoft Exchange Load Generator tool.
3. Open the saved configuration file in Notepad or a similar application, manually change the
required task parameters, and then save the file when you are completed.
4. Open the Microsoft Exchange Load Generator tool. (Locate LoadGenWin.exe and double-
click it.)
5. On the Welcome to Exchange Load Generator landing page, click Start a new test,
provide the saved configuration file name on the next page, and then click Continue.
6. Continue paging through until you reach the Configuration summary page, and then click
Start the initialization phase (recommended before running the test).
Changing the XML

The following are the tasks that can be run during simulation. To change the XML, search for the
task and change the parameters to customize the file. See Customizing the Exchange Load
Generator Test for information about the parameters.
 BrowseCalendar
42
 BrowseContacts
 BrowsePublicFolder
 BrowseTasks
 CreateContact
 CreateFolder
 CreateTask
 DeleteFolder
 DeleteMail
 DownloadOab
 EditRules
 EditSmartFolders
 ExportMail
 Logoff
 Logon
 MailboxStoreProfile
 MakeAppointment
 MoveMail
 PostFreeBusy
 PublicFolderPost
 PublicStoreProfile
 ReadAndProcessMessages
 RequestMeeting
 SendMail
 Search
Using System Monitor and Exchange Load

Generator Data
Exchange Load Generator performance counters represent only a partial view of the operation of
a system. To fully understand how the system works, analyze the information in the Windows and
Exchange counters by running Windows® System Monitor during the Exchange Load Generator
test.
Use System Monitor data to do the following:
 Track how closely various resources are to saturation, and adjust subsequent Exchange
Load Generator tests accordingly.
 Use the server performance data to identify bottlenecks and resources that are in high
demand, and consider re-tuning the system and then running the same test again.
43
You should log the System Monitor data from both the Exchange server and the Exchange Load
Generator client on a single, separate computer. For more information about consolidating your
log files on a single computer, see How to Consolidate System Monitor Data.
Questions to Consider when Looking at Test Results

Consider the following questions when looking at Exchange Load Generator test results and the
performance data from System Monitor. To additionally help you answer these questions, view
the log files in Exchange Load Generator and System Monitor.
 Were there any errors in the test? Did they significantly affect the load?
You can use both the LoadGen.log file and Windows Event Viewer to locate any test errors.
To determine the validity of the test, examine the type and quantity of errors.
 Did the client computers create the expected load?
If too many users are simulated per client computer, Exchange Load Generator generates
less than the expected load. To determine whether too many users are causing Exchange
Load Generator to generate a lighter load, you can designate a control client that has one-
fourth or one-eighth the number of users as the other client computers as a control for your
test. If the per-user rate of events that Exchange Load Generator generates is different for the
control client and the other clients, your experiment is invalid because the workload that is
generated is incorrect. If your other client computers are bottlenecked, they will generate less
than the expected requests. Additionally, if you compare the response times between the
regular clients and the control client, you can help determine whether the number of users per
client computer is affecting the results.
 What was the actual load on the server with regard to transactions per user per day? Does
this match the expected value?
Make sure that you answer these two questions. To better understand the accuracy of the
test and the performance characteristics of a server, consider the actual work that the server
performed. If the server performed a different amount of work than expected, your test
represents a different workload than intended.
 Were there any server resources that were saturated? Were there any software queues
growing unbounded?
To verify this we suggest you use the performance counters and thresholds that are specified
in the following locations.
Exchange 2010
http://technet.microsoft.com/en-us/library/dd335215.aspx
How to Consolidate System Monitor Data

You can consolidate System Monitor (perfmon) data by logging all Exchange Load Generator
client counters and Exchange server System Monitor counters to one data log instead of having
one log per computer.
44
To log all Exchange Load Generator counters and Exchange server counters to one
data log
1. Log on to the computer on the domain where you want to log all System Monitor data.
2. Open Services: Click Start, point to Programs, point to Administrative Tools, click
Computer Management, and then click Services.
3. In Services, in the results pane, under Name, right-click Performance Logs and Alerts,
and then click Properties.
4. In Performance Logs and Alerts Properties (Local Computer), click the Log On tab,
and then click This account.
5. Click Browse.
6. In Select User, select the domain administrator account, and then click OK.
7. On the Log On tab, in the Password box, type the password for the domain
administrator account.
8. In the Confirm password box, retype the password for the domain administrator
account.
9. Click OK.
10. Close Services.
11. Click Start, click Run, and then type perfmon.msc to access the Performance dialog
box.
12. In Performance, click System Monitor, and then click the Add (+) icon on the toolbar to
open the Add Counters dialog box.
13. In the Add Counters dialog box, click Select counters from computer, and then type
the server or client name in the corresponding list.
14. In the Performance object list, select the performance object to add. Minimally, you
should add Exchange Load Generator Engine and Exchange Load Generator User
Groups for each client.
15. Click All counters to add all counters for the object, or click Select counters from list to
individually select the counters to be added for the object.
16. Repeat Steps 14 and 15 for each performance object to be monitored for this computer.
17. Repeat Steps 12 to 16 for each additional client or server for which you want to
consolidate the logs.
Operational Recommendations for Using

Exchange Load Generator
The following operational recommendations will help you achieve solid results from the Exchange
Load Generator tests.
45
Simulating Cached Mode
Exchange Load Generator simulates the cached mode of Microsoft Office Outlook 2003 and later
versions, through a combination of tasks that are common to online mode. To simulate cached
mode, use the Cached Mode client type as a starting point to increasing and decreasing task
frequencies to match your user profiles.
Using Stress Mode

By using stress mode, you can run a test with the Exchange Load Generator users performing
their tasks as fast as they can. When in this mode, Exchange Load Generator does not spread
out the tasks throughout the defined length of a simulation day, but runs them as fast as the
server or client can and will finish probably faster and unpredictably.
You should only use this mode when the objective is to stress the server at its mode and ignore
the deterministic way (spreading a fixed load for a fixed number of hours) that Exchange Load
Generator runs. Make sure while running in stress mode that you are not bound by the
performance of your Loadgen clients but of the server.
RPC over HTTP

To use RPC over HTTP, your topology has to be configured to use one or more RPC proxy
servers. Instructions for doing this using the configuration file are in the topic, Customizing the
Exchange Load Generator Test. The following are two recommendations for user loads when you
use RPC over HTTP.
Does the Connector Behavior Accurately Simulate Outlook?

With a reasonable number of users who are running on an Exchange Load Generator computer,
inactive RPC/HTTP connections will be disconnected automatically by the Windows RPC runtime
component at 90-second intervals. This means that with such a small level of load running
through RPC/HTTP, you may find that the number of active connections is very low.
How Many Users does the Client Support?

The number of users who are supported by the client depends on the actions and frequencies
specified in the configuration file. Generally, the number of users is limited by the connections and
ports configured on the server. If you are running Outlook 2003 online simulation, there are two
RPC connections to the mailbox store and one to the public folder for each user. Each connection
generates two HTTP connections (data in and data out). Make sure that the server is configured
correctly and there are no other applications on either side that consume more TCP port numbers
than you expected. For a client computer that has a single processor (2 GHz) and 1 gigabyte of
RAM, no major issues are presented in supporting 1200 to 1500 RPC over HTTP users driven by
Exchange Load Generator. The command-generated LoadGenConfig.xml file would be used for
setting up the configuration.
Make sure that you monitor the value of \Exchange Load Generator Engine\Task Queue Length
to make sure that it is not increasing. The value of \Exchange Load Generator Engine\Task
46
Completed/sec and \Exchange Load Generator Engine\Task Dispatched/sec should also match.
This means that the client is catching up. Usually the out-of-memory problem or MAPI connection
errors occur you hit the CPU bottleneck.
Distribution Groups, Query-Based Distribution Groups, and

Global Catalogs
In the Read and Process Message task, you can configure a percentage of the messages to be
sent to distribution groups (also known as DLs) or query-based distribution groups. Although you
can configure the average size of a distribution group in Exchange Load Generator, query-based
distribution groups are generally larger than the distribution groups. This is because they can
include all per database, per server, or for a whole topology.
Sending mail to query-based distribution groups can severely stress the server and global
catalogs’ domain controller when the query-based distribution groups are being expanded. We
recommend that you set up a dedicated expansion server for the distribution groups. You should
also modify Exchange Load Generator’s distribution groups and query-based distribution groups
accordingly. Additionally, consider using the “one per MDB” query-based distribution group that is
typically smaller.
Deployment Verification Using Exchange Load Generator

Exchange Load Generator can be an effective tool for deployment verification. For example, you
can use it to verify that you will be able to support the number of users who have been planned
for your deployment. You can also examine how the system might react to stress loads that are
heavier than what you would typically expect.
Important:
Although Exchange Load Generator can be used to verify your deployment in various
ways, it should not be used in production environments.
We recommend that Exchange Load Generator be run for verification in separate topologies and
then rolled out on production environments. Also, it is a good idea to run each Exchange Load
Generator test multiple times. This will help make sure of the validity of your results.
Capacity Planning Using Exchange Load Generator

Exchange Load Generator has profiles that enable you to verify deployments and that help in
capacity planning. To model your deployment’s users, enable or disable tasks and change the
settings and task frequencies of the 200 profile. The 200 profile has many features to better
simulate real-life users.
Caution:
We encourage you to use the well-known profiles such as "200" that are provided by
Exchange Load Generator. Avoid creating a custom profile unless you are very familiar
with Exchange Load Generator parameters, and you understand what type of difference
your customization will make in terms of the load that is generated on the server.
47
To control the stress load that Exchange Load Generator puts on the Exchange server, you can
change the duration settings of the simulation in the Specify test settings page of the graphical
user interface, and enable or disable tasks. Exchange Load Generator calculates the stress from
the frequencies of the tasks, and then spreads the load homogeneously across the configured
length of the simulation day. For example, if Exchange Load Generator must run a task four times
in an eight-hour daytime period, it will run a task every two hours. By increasing the task
frequency to eight times a day and keeping the length of the simulation day setting the same, a
task will be run every hour, thereby doubling the stress load. Therefore, if the length of the
simulation day increased, the stress is actually diminished, because the same load is spread over
a longer time. To increase the stress load, increase the task frequencies and/or reduce the
simulation day length.
Re-create the Topology Every Time a Topology Setting Changes

After you change the topology settings, you should re-create the topology to make sure that the
physical topology matches your settings.
Back up and Restore Databases to Skip Repeating the

Initialization Phase
After you initialize users for a test run that will be changed and repeated over time, it is a good
idea to back up the databases so that the initialization process does not have to be repeated.
After each test run, restore the databases and re-start your tests without having to reinitialize all
users.
Note:
Backing up and restoring databases to skip repeating the initialization phase makes
sense only if the initialization and topology settings are unchanged between test runs for
the server in question.
Monitor for Saturation

The typical goal of Exchange Load Generator tests is to analyze the load that the server can
support and still provide sufficient response times. However, sometimes the goal of Exchange
Load Generator tests is to determine what load the network can support while still providing
sufficient response times. You should saturate the resource you are focusing on with the test, but
not saturate other resources in the test. For example, you cannot effectively study the server
behavior if the client computers or the network are saturated.
The following two methods can be used to monitor for saturation:
 Monitor the demands on the clients, network, and server.
 Do not let the clients become the bottleneck in an experiment.
Monitor Mail Queues

Use System Monitor to make sure that mail delivery is working correctly.
48
If there are several messages in the queue, this indicates a server bottleneck. Because the server
must operate at a consistent baseline for sufficient benchmarking, a server bottleneck invalidates
your test results. For example, if the queue is continually growing for the whole duration of the
test, the server is probably overloaded. This indicates a bottleneck.
Specific Features and Behaviors of Exchange

Load Generator
This topic contains in-depth observations about some specific features and behaviors of
Exchange Load Generator.
Public Folders
Exchange Load Generator has two public folder tasks: Public Folder Post and Browse Public
Folder. Both tasks will choose any public folder in the MAPI TLH (the default public folder root),
except for one named Internet Newsgroups.
Exchange Load Generator can choose other folders to browse or post, even if it did not create
them. Therefore, in such an attempt, Exchange Load Generator may not always run some tasks
and indicate it had access denied. We recommend that you have only the public folders that were
created by Exchange Load Generator.
When Exchange Load Generator creates its public folders, it changes the permissions to give
owner rights as default. This means that any user can edit, delete, or add new posts. This is
required for Exchange Load Generator to simulate multiple users when logging on to users using
their respective accounts. If the permissions are changed in any way, Exchange Load Generator
might not always create posts or browse the public folders.
Mail Flow and Recipients

A core functionality of Exchange Load Generator is that it can send mail to users, distribution
groups, and query-based distribution groups. When you set up a complex topology, you must
understand how Exchange Load Generator will decide who it will send mail to.
There are two distinct scenarios that cause Exchange Load Generator to send mail. The first is
when Exchange Load Generator is processing the Inbox or browsing old mail and tries to Reply
or Reply all to one or more messages. In this scenario, Exchange Load Generator sends the
reply to the sender and recipients (if Reply all), regardless of whether the users are in the
topology. This means that when a topology is initialized, all the mail that is populated in the
mailbox is consistent with the topology settings of the client that initialized all mailboxes. As test
runs occur, e-mail is added from clients that have their own topology configuration, which can be
the same as the one that initialized the mail. If at any point there is a change in the topology (for
example, a server is removed or users are added or removed), the mail that is already in the
user’s mailboxes is not affected. If Exchange Load Generator tries to reply to a sender that no
longer exists, a non-delivery report (NDR) occurs.
49
The other scenario is with regard to sending a new mail. When sending a new mail, Exchange
Load Generator uses the information in the send mail task for selecting the number of recipients
and whether distribution groups and query-based distribution groups should be added. After this
step, Exchange Load Generator chooses the recipients based on the test group configuration.
For example, if the topology settings are configured so that a server has 1,000 users and the test
targets only 10 users, Exchange Load Generator will only consider 10 users as possible
recipients. Additionally, the distribution groups and query-based distribution groups will be
configured to the 1,000 users. Therefore any mail going to them will end up in users other than
the 10 users.
Task Rate, Concurrency Settings, and Stress Mode

To fine-tune stress, it helps to understand how Exchange Load Generator calculates how much
load will occur against servers. This calculation process occurs in two steps.
In the first step, Exchange Load Generator calculates the number of tasks that need to run per
user group. This is the sum of all task frequencies multiplied by the number of users.
In the second step, Exchange Load Generator sums the number of tasks of all the user groups
that the client computer that the test is running on is configured to do. In other words, this is the
result of the calculation from the first step. The summed task is spread across the total duration of
the simulation. The duration values are set in the Specify test settings page in the graphical
user interface as follows:
 The length of a 'simulation day' defined in hours, minutes, and seconds
 The total length of the simulation defined as forever, or as a specified duration (in hours,
minutes, and seconds)
For example, if there is one group that has 100 users and will run the Send Mail task 10 times in
an eight-hour configuration, Exchange Load Generator will spread 1,000 tasks across eight
hours, running each task every 347 milliseconds (1,000 divided by 8 hours).
Independently of the concurrency settings, the duration of the simulation will stop the test after a
designated elapsed time is started. For the previous example, if the duration of the simulation
was two or eight hours, the tasks would still run every 347 milliseconds. The difference is that the
test would stop in two and eight hours, respectively.
Finally, there is an additional option: stress mode. In stress mode, Exchange Load Generator
ignores the concurrency settings and task frequencies. It runs the tasks back-to-back and without
a wait for the duration of the simulation. For example, if the duration of the simulation was set to
eight hours, the tasks would run for eight hours without the 347 millisecond wait between them.
Because only the client, server, and network configurations limit this mode, results will vary
extensively from setup to setup. Therefore, we do not recommend this mode for comparing
computers or for sizing.
50
Dynamic Mail Generator
Mail messages are now created dynamically as they are needed. This ensures a more unique
message mix that is closer to a real environment. The word frequency it based on the language of
choice.
The dynamic mail generator has a number of settings that you can specify.
o Min/Max/Avg message size
o Default language (English, French, Spanish, Korean, Turkish)
o Language Variation %
o Default Body Type (html, rtf, text)
o Body Type Variation %
o Attachments (just drop the attachments that you want to use into the folder where
the Dynamic Mail Generator is looking.
The dynamic mail generator is setup by default for initialization as well as sending mail.
We currently do not have a UI for this feature. You will have to customize your Loadgen XML to
change the default settings.
The default settings are:

<GeneratedMailProfile>
<TaskConfigList>
<TaskConfig d6p1:type="GeneratedMailProfileTask"
<TaskParameters d6p1:type="GeneratedMailParameters">
<MessageSizeParameters>
<MinimumSize>100</MinimumSize>
<MaximumSize>100000</MaximumSize>
<AverageSize>30000</AverageSize>
<IndexableTextPercentage>100</IndexableTextPercentage>
</MessageSizeParameters>
<LanguageParameters>
<EnglishLanguageWeight>100</EnglishLanguageWeight>
<FrenchLanguageWeight>0</FrenchLanguageWeight>
<KoreanLanguageWeight>0</KoreanLanguageWeight>
<SpanishLanguageWeight>0</SpanishLanguageWeight>
<TurkishLanguageWeight>0</TurkishLanguageWeight>
</LanguageParameters>
<BodyTypeParameters>
<HtmlBodyWeight>70</HtmlBodyWeight>
<PlainTextBodyWeight>5</PlainTextBodyWeight>
<RtfBodyWeight>15</RtfBodyWeight>
</BodyTypeParameters>
<AttachmentParameters>
<AttachFrequency>0.05</AttachFrequency>
<AttachmentCount>
<MinimumCount>1</MinimumCount>
<MaximumCount>5</MaximumCount>
<AverageCount>1</AverageCount>
</AttachmentCount>
<PercentInfected>0</PercentInfected>
<MessageFiles>
<MessageFile>
<Weight>840</Weight>
<Pathname>messages\msg00001.msg</Pathname>
</MessageFile>
</MessageFiles>
</AttachmentParameters>
</TaskParameters>
</TaskConfig>
</TaskConfigList>
</GeneratedMailProfile>
51
In order to use attachments, specify them in the MessageFiles section of the
GeneratedMailProfile.
Using Custom Messages

Exchange Load Generator has two areas where a custom message mix can be used. One set is
for sending new messages and another set is for posting in public folders. The former is defined
in the Send Mail task and it is also used by individual mailbox initialization. The latter is defined in
the Public Folder Post task.
Exchange Load Generator has one message set for public store initialization. It is also configured
in the LoadGenConfig.xml file.
You can use any MSG MAPI message that is formatted correctly with Exchange Load Generator.
The following table lists the message mix of the files that are downloaded to the installation
directory. The attachments can be documents, clip art, spreadsheets, or any type recognized by
the client mail application.
Exchange Load Generator loads in memory all messages used from both message sets, per user
group. This means that if you have multiple user groups, even if with the same list of messages to
be used, they will be loaded separately and occupy more space in memory. Additionally, the
overall memory consumption of Exchange Load Generator increases if many custom messages
are used.
The following table lists the message mix of the files that are downloaded to the installation
directory. Use these files as is, or customize for your test simulation.
File name Content type Size (in bytes) in DAV
msg00001 HTML 2187
msg00002 HTML 3858
msg00003 HTML 5174
msg00004 HTML 6716
msg00005 HTML 11639
msg00006 HTML 16720
msg00007 HTML 32518
msg00008 HTML 47781
msg00009 HTML 105244
msg00010 HTML 211710
msg00011 HTML 313025
msg00012 HTML 530583
msg00013 HTML 106890
52
msg00014 HTML 2651291
msg00015 HTML w/attachment 6831
msg00022 RTF 6404
msg00023 RTF 8280
msg00024 RTF 10174
msg00025 RTF 20850
msg00026 RTF 30254
msg00027 RTF 61812
msg00028 RTF 92586
msg00029 RTF 205005
msg00030 RTF 409874
msg00031 RTF 614404
msg00032 RTF 1048846
msg00033 RTF w/attachment 12217
msg00040 Text 1472
msg00041 Text 2287
msg00042 Text 2681
53
msg00043 Text 3190
msg00044 Text 4900
msg00045 Text 6624
msg00046 Text 11679
msg00047 Text 16607
msg00048 Text 35177
msg00049 Text 68834
msg00050 Text 102574
msg00051 Text 346338
msg00052 Text w/attachment 5661
Performance Counters Reference

You can monitor the progress of the simulation when the test is being run by observing
performance counters. To reach the Performance screen, click Start, click Run, type perfmon,
and then click OK.
The following two tables list the counters and purposes for the Exchange Load Generator
Engine Counter object and the Exchange Load Generator User Groups Counter object.
Note:
Do not use the Exchange Load Generator latency counters as a measure of acceptable
performance with regard to latency. Instead, use the RPC averaged latency counter on
the server.
Counter Object for Exchange Load Generator Engine
54
Counter Purpose
ContentCacheTrims Number of times that the task engine was

required to allocate additional space for cached
data, where there were no more than 8 regions
of 64 MBs available and in-memory contents
were removed from the cache. Indicates that
task engine does not have sufficient virtual
memory to perform an accurate simulation.
ContentCacheTrims/sec. Rate of content cache trims. As the rate

increases, overall accuracy of the simulation
decreases.
Task Exceptions Count of task executions that resulted in a fatal

exception
Task Exceptions/sec. Number of fatal exceptions encountered when

running a task per second
Task Interval Number of milliseconds that the engine will wait

between dispatching tasks
Task Queue Length Number of items currently in task queue that

have not yet been executed
Tasks Skipped Number of tasks that were skipped
Tasks Completed Total number of tasks completed
Tasks Completed/sec. Number of tasks completed per second
Tasks Dispatched Total number of tasks dispatched
Tasks Dispatched/sec. Number of tasks dispatched per second
Counter Object for Exchange Load Generator User Groups

The counters are listed alphabetically, and within task or action type.
Counter Purpose
Browse Calendar Action Latency Average time (in milliseconds) taken to

complete a Browse Calendar action
Browse Calendar Actions Total Browse Calendar actions executed
Browse Calendar Actions/sec. Count of Browse Calendar actions executed

per second
Browse Calendar Action Latency (95th The 95th percentile value of the time (in
55
Counter Purpose
Percentile) milliseconds) taken to complete a Browse

Calendar action
Browse Contacts Action Latency Average time (in milliseconds) taken to

complete a Browse Contacts action
Browse Contacts Actions Total Browse Contacts actions executed
Browse Contacts Actions/sec. Count of Browse Contacts actions executed per

second
Browse Contacts Action Latency (95th The 95th percentile value of the time (in
Contacts action
Browse Public Folder Action Latency Average time (in milliseconds) taken to
complete a Browse Public Folder action
Browse Public Folder Actions Total Browse Public Folder actions executed
Browse Public Folder Actions/sec. Count of Browse Public Folder actions

executed per second
Browse Public Folder Action Latency (95th The 95th percentile value of the time (in
Public Folder action
Browse Tasks Action Latency Average time (in milliseconds) taken to

complete a Browse Tasks action
Browse Tasks Actions Total Browse Tasks actions executed
Browse Tasks Actions/sec. Count of Browse Tasks actions executed per

second
Browse Tasks Action Latency (95th Percentile) The 95th percentile value of the time (in
milliseconds) taken to complete a Browse
Tasks action
Create Contact Action Latency Average time (in milliseconds) taken to

complete a Create Contact action
Create Contact Actions Total Create Contact actions executed
Create Contact Actions/sec. Count of Create Contact actions executed per

second
Create Contact Action Latency (95th Percentile) The 95th percentile value of the time (in
milliseconds) taken to complete a Create
Content action
56
Counter Purpose
Create Folder Action Latency Average time (in milliseconds) taken to

complete a Create Folder action
Create Folder Actions Total Create Folder actions executed
Create Folder Actions/sec. Count of Create Folder actions executed per

second
Create Folder Action Latency (95th Percentile) The 95th percentile value of the time (in
milliseconds) taken to complete a Create
Folder action
Create Task Action Latency Average time (in milliseconds) taken to

complete a Create Task action
Create Task Actions Total Create Task actions executed
Create Task Actions/sec. Count of Create Task actions executed per

second
Create Task Action Latency (95th Percentile) The 95th percentile value of the time (in
milliseconds) taken to complete a Create Task
action
Delete Mail Action Latency Average time (in milliseconds) taken to

complete a Delete Mail action
Delete Mail Actions Total Delete Mail actions executed
Delete Mail Actions/sec. Count of Delete Mail actions executed per

second
Delete Mail Action Latency (95th Percentile) The 95th percentile value of the time (in
milliseconds) taken to complete a Delete Mail
action
Download OAB Action Latency Average time (in milliseconds) taken to

complete a Download OAB action
Download OAB Actions Total Download OAB actions executed
Download OAB Actions/sec. Count of Download OAB actions executed per

second
Download OAB Action Latency (95th The 95th percentile value of the time (in
Percentile) milliseconds) taken to complete a Download
OAB action
Edit Rules Action Latency Average time (in milliseconds) taken to

complete an Edit Rules action
57
Counter Purpose
Edit Rules Actions Total Edit Rules actions executed
Edit Rules Actions/sec. Count of Edit Rules actions executed per

second
Edit Rules Action Latency (95th Percentile) The 95th percentile value of the time (in
milliseconds) taken to complete an Edit Rules
action
Edit Smart Folders Action Latency Average time (in milliseconds) taken to
complete an Edit Smart Folders action
Edit Smart Folders Actions Total Edit Smart Folders actions executed
Edit Smart Folders Actions/sec. Count of Edit Smart Folders actions executed
per second
Edit Smart Folders Action Latency (95th The 95th percentile value of the time (in
Percentile) milliseconds) taken to complete an Edit Smart
Folders action
Export Mail Action Latency Average time (in milliseconds) taken to

complete a Export Mail action
Export Mail Actions Total Export Mail actions executed
Export Mail Actions/sec. Count of Export Mail actions executed per

second
Export Mail Action Latency (95th Percentile) The 95th percentile value of the time (in
milliseconds) taken to complete an Export Mail
action
Initialize Mailbox Action Latency Average time (in milliseconds) taken to

complete an Initialize Mailbox action
Initialize Mailbox Actions Total Initialize Mailbox actions executed
Initialize Mailbox Actions/sec. Count of Initialize Mailbox actions executed per

second
Initialize Mailbox Action Latency (95th The 95th percentile value of the time (in
Percentile) milliseconds) taken to complete an Initialize
Mailbox action
Logoff Action Latency Average time (in milliseconds) taken to

complete a Logoff action
Logoff Actions Total Logoff actions executed
Logoff Actions/sec. Count of Logoff actions executed per second
58
Counter Purpose
Logoff Action Latency (95th Percentile) The 95th percentile value of the time (in
milliseconds) taken to complete a Logoff action
Logon Action Latency Average time (in milliseconds) taken to

complete a Logon action
Logon Actions Total Logon actions executed
Logon Actions/sec. Count of Logon actions executed per second
Logon Action Latency (95th Percentile) The 95th percentile value of the time (in
milliseconds) taken to complete a Logon action
Make Appointment Action Latency Average time (in milliseconds) taken to

complete a Make Appointment action
Make Appointment Actions Total Make Appointment actions executed
Make Appointment Actions/sec. Count of Make Appointment actions executed

per second
Make Appointment Action Latency (95th The 95th percentile value of the time (in
Percentile) milliseconds) taken to complete a Make
Appointment action
Move Mail Action Latency Average time (in milliseconds) taken to

complete a Move Mail action
Move Mail Actions Total Move Mail actions executed
Move Mail Actions/sec. Count of Move Mail actions executed per

second
Move Mail Action Latency (95th Percentile) The 95th percentile value of the time (in
milliseconds) taken to complete a Move Mail
action
Not Implemented Action Latency Average time (in milliseconds) taken to

complete a Not Implemented action
Not Implemented Actions Total Not Implemented actions executed
Not Implemented Actions/sec. Count of Not Implemented actions executed

per second
Not Implemented Action Latency (95th The 95th percentile value of the time (in
Percentile) milliseconds) taken to complete a Not
Implemented action
Post Free Busy Action Latency Average time (in milliseconds) taken to
complete a Post Free Busy action
59
Counter Purpose
Post Free Busy Actions Total Post Free Busy actions executed
Post Free Busy Actions/sec. Count of Post Free Busy actions executed per
second
Post Free Busy Action Latency (95th The 95th percentile value of the time (in
Percentile) milliseconds) taken to complete a Post Free
Busy action
Public Folder Post Action Latency Average time (in milliseconds) taken to
complete a Public Folder Post action
Public Folder Post Actions Total Public Folder Post actions executed
Public Folder Post Actions/sec. Count of Public Folder Post actions executed
per second
Public Folder Post Action Latency (95th The 95th percentile value of the time (in
Percentile) milliseconds) taken to complete a Public Folder
Post action
Read And Process Messages Action Latency Average time (in milliseconds) taken to
complete a Read And Process Messages
action
Read And Process Messages Actions Total Read And Process Messages actions
executed
Read And Process Messages Actions/sec. Count of Read And Process Messages actions
executed per second
Read and Process Action Latency (95th The 95th percentile value of the time (in
Percentile) milliseconds) taken to complete a Read and
Process action
Request Meeting Action Latency Average time (in milliseconds) taken to

complete a Request Meeting action
Request Meeting Actions Total Request Meeting actions executed
Request Meeting Actions/sec. Count of Request Meeting actions executed per

second
Request Meeting Action Latency (95th The 95th percentile value of the time (in
Percentile) milliseconds) taken to complete a Request
Meeting action
Search Action Latency Average time (in milliseconds) taken to

complete a Search action
60
Counter Purpose
Search Actions Total Search actions executed
Search Actions/sec. Count of Search actions executed per second
Search Action Latency (95th Percentile) The 95th percentile value of the time (in
milliseconds) taken to complete a Search
action
Send Mail Action Latency Average time (in milliseconds) taken to

complete a Send Mail action
Send Mail Actions Total Send Mail actions executed
Send Mail Actions/sec. Count of Send Mail actions executed per

second
Send Mail Action Latency (95th Percentile) The 95th percentile value of the time (in
milliseconds) taken to complete a Send Mail
action
Synchronize Folders Action Latency Average time (in milliseconds) taken to

complete a Synchronize Folders action
Synchronize Folders Actions Total Synchronize Folders actions executed
Synchronize Folders Actions/sec. Count of Synchronize Folders actions executed

per second
Synchronize Folders Action Latency (95th The 95th percentile value of the time (in
Percentile) milliseconds) taken to complete a Synchronize
Folders action
Task Exceptions Count of task executions that resulted in a fatal

exception within the current user group
Task Exceptions/sec. Count of task executions that resulted in a fatal

exception per second within the current user
group
User Init Action Latency Average time (in milliseconds) taken to

complete a User Initiation action
User Init Actions Total User Initiated actions executed
User Init Actions/sec. Count of User Initiated actions executed per

second
User Init Action Latency (95th Percentile) The 95th percentile value of the time (in
milliseconds) taken to complete a User Init
action
61
Counter Purpose
User Term Action Latency Average time (in milliseconds) taken to

complete a User Term action
User Term Actions Total User Term actions executed
User Term Actions/sec. Count of User Term actions executed per

second
User Term Action Latency The 95th percentile value of the time (in
milliseconds) taken to complete a User Term
action
Users Active Number of active users in the current

configuration
Users Configured Number of users defined in the current

configuration
Users Standby Number of inactive users in the current

configuration
Scripted Load Modules in the Exchange Load

Generator Framework
The following topics describe the scripted load modules in the Exchange Load Generator
framework.
 Description and example of the Exchange Load Generator configuration XML at Exchange
Load Generator Configuration XML.
 Description of the base set of verbs that all the modules can use at Base Script Verbs.
 Information about the POP3 module at The POP3 Module.
 Information about the SMTP module at The SMTP Module.
 Information about the IMAP4 module at The IMAP4 Module.
 Information about the Outlook Web Access module at The Outlook Web Access Module.
Exchange Load Generator Configuration XML

The Exchange Load Generator configuration XML is used to supply arguments to the script
modules. In particular, <ScriptProfile> defines the scripted module-specific parameters.
The following table explains the tags that are used in the <ScriptProfile>.
Tag Description
Logging True or false.
62
Tag Description
LogPath String. The default is the executing directory

and module name.
LogLevel Integer. 0-2 increasing amount of detail in the

log.
ScriptPath String. Path to the script file.
EncryptedPort True or false. By default, the target port is

encrypted.
Server String. The target or destination server.
Port Integer. Port to which requests will be directed.
ProtocolName String. The name of the protocol.
StartDelay Integer. The interval (in) milliseconds to wait

between the start of each user instance.
UserCount Integer. Number of script instances. (-1 will run

entire UserGroup.)
NearBy Integer. Distance from the current user to

search for other users.
Sample Configuration XML

The XML script is as follows:
<UserGroup>
<ExtendedProperties />
<ScriptProfile>
<ExtendedProperties />
<Logging>true</Logging>
<LogPath>\myModuleLog.txt</LogPath>
<ScriptPath>myModuleScript.txt</ScriptPath>
<EncryptedPort>false</EncryptedPort>
<Server>targetServer</Server>
<Port>25</Port>
<StartDelay>4</StartDelay>
<UserCount>30</UserCount>
<NearBy>1</NearBy>
</ScriptProfile>
63
<ClientType>SmtpModule</ClientType>
<Name>UserGroup2</Name>
<LdapBaseDn>OU=FewUsers,DC=XZPVCB,DC=com</LdapBaseDn>
<StartWaitDuration>P0Y0M0DT0H0M0S</StartWaitDuration>
<RunLengthDuration>P0Y0M0DT14H0M00S</RunLengthDuration>
</UserGroup>
<ClientType/>under<UserGroup/> is used to define which scripted module to simulate. Following
is a list of client types supported by Exchange Load Generator.
<ClientType>SmtpModule</ClientType>
<ClientType>PopModule</ClientType>
<ClientType>ImapModule</ClientType>
<ClientType>Owa2013Module</ClientType>
Base Script Verbs

The base class provides a set of verbs that can be used by all the modules. The script can also
contain variables that are described in this topic.
Note:
Instances of variables and verbs are not related. That is, SEQULIST will be sequential for
each script instance.
Table of Verbs
The following table lists the verbs and their functions.
Verb Function
RANDLIST Returns a string from the list of strings inside a

file.
Matches "randlist\$([-_:\\\\.#a-z0-9A-Z]+)\$"
Example: RANDLIST(c:\temp\myfilelist.txt)
returns “hello14”
SEQULIST Returns the next string from a list of strings.

Wraps after the ultimate line.
Matches "sequlist\$([-_:\\\\.#a-z0-9A-Z]+)\$"
Example: SEQULIST(c:\temp\myfilelist.txt)
SAMESEQULIST Returns the same string as the previous call to

SEQULIST.
64
Verb Function
Matches "samesequlist\$([-_:\\\\.#a-z0-9A-Z]
+)\$"
Example: SEQULIST(C:\temp\myfilelist.txt)
Example: SAMESEQULIST(C:\temp\
myfilelist.txt) returns “hello0”
SEQUNUMBER Returns the next number between n1 and n2.

Matches "sequnumber\$([0-9,]+\$)"
Example: SEQUNUMBER(0,10) returns 1
SAMESEQUNUMBER Returns the same number as the previous call

to SEQUNUMBER.
Matches "samesequnumber\$([0-9,]+\$)"
Example: SAMESEQUNUMBER(0,10) returns
1
RANDNUMBER Returns a random number between n1 and n2.

Matches "randnumber\$([0-9,]+\$)"
Example: RANDNUMBER(0,100) returns 32
LOOP Starts a loop block.

Matches "loop ([0-9]+)"
Example: LOOP (10)
ENDLOOP Ends a loop block.

Matches "endloop"
Example: ENDLOOP
RANDNUMERIC Returns a random number of n digits.

Matches "randnumeric\$([0-9]+)\$"
Example: RANDNUMERIC(4) returns 3456
RANDALPHA Returns a random string of length n.

Matches "randalpha\$([0-9]+)\$"
Example: RANDALPHA(3) returns sdk
SKIP Skips the next n lines, wraps, not including lines

that have been removed or blanks.
Matches "skip\$([0-9]+)\$"
Example: SKIP(10)
PERCENT Executes the remainder of the line n% of the
65
Verb Function
time.
Matches "%([0-9]+)"
Example: %40 RCPT TO: [email protected]
SLEEP Pauses execution of the instance in

milliseconds.
Matches "sleep"
Example: SLEEP 1000
RANDUSER Selects a random user from the list of users in

the UserGroup.
Information available: username, password, e-
mail, domain.
Matches "RANDUSER.[a-zA-Z]+"
Example: RANDUSER.emailaddress returns
[email protected]
SEQUUSER Selects the next user from the list of users in

the UserGroup.
mail, domain
Matches "SEQUUSER.[a-zA-Z]+"
Example: SEQUUSER.username returns
user21
NEARBY Randomly picks a user from the list of users in

the UserGroup within n of the current user.
ScriptProfile xml parameter NearBy default
value = 5
mail, domain
Matches "NEARBY.[a-zA-Z]+"
Example: NEARBY.emailaddress returns
[email protected]
USER Replaced with the current users information.

This is a one-time replacement, not a variable.
mail, domain
Matches “USER.[a-zA-Z]”
USER.emailAddress
66
Verb Function
USER.username
USER.userPassword
USER.domain
Variable Behavior
All script-based modules support simple variables which enable you to store values and do
simple operations. Declaration and initial value is:
 The default value is string.empty.
 There are 2 types: string and integer.
 There are no operators in the SET command.
The reserved patterns are "\$[^\$^ ]+\$" and "~userVar".
SET <$var$> [initial value]
SET $myvar$
SET $myvar1$ 1
SET $myVAR2$ the world
SET $MyVar3$ RANDNUMBER(0,1)
NO SET $MyVar4$ $Myvar3$
NO RANDNUMBER(0,$myVar$)
Integer Operators
All operations return a string: += and -=. Change the value of the left operand. Operations other
than SET on a string variable return the unchanged value of the variable. Precedence is right to
left replacement with the universal variables being replaced first.
+
$myvar$ + 4
$myVar$ + $myvar2$
$myVar$ + RANDNUMBER(0,100)
++
$myvar$++
+=
$myvar$ += 5
$myvar$+=$myvar2$
-
$myvar$ - 4
$myVar$ - $myvar2$
67
$myVar$ - RANDNUMBER(0,100)
--
$myvar$--
-=
$myvar$ -= 5
$myvar$ -= $myvar2$
The POP3 Module

The Post Office Protocol version 3 (POP3) module simulates multiple clients using the POP3
protocol. The module supports parameters to configure options for the following:
 Different protocol-specific options
 How the test executes
 Security options
Examples are the server that this client is targeting or the port number. Client actions are
simulated by running a script that contains scripted module keywords and POP3 verbs.
This module also implements performance counters for protocol statistics. Additionally, you can
log client actions and server responses during a test run. You cannot configure the events that
are logged.
Script Verbs Specific to the POP3 Module

The following table lists script verbs and their functions.
Script verb Function
AUTH Authenticates the current user.
CAPA Executes the CAPA request.
CONNECT Connects to the specified server on the

specified port.
DELE <msgnum> Marks the specified message for deletion.
LIST Returns a listing of messages by using the

current message number and size.
LOGON Creates a logon request using the current user.
NOOP Sends a NOOP (no operation) request.
QUIT Ends the conversation.
RETR <msgnum> Retrieves the specified message.
RSET Resets the status on the mailbox. All messages

that are marked for deletion will be unmarked.
68
STAT Returns the number of items in the mailbox and

the total size of the mailbox.
STLS Starts a TLS session.
TOP <msgnum> Retrieves the requested number of lines from

the specified message.
UIDL <msgnum> Returns a Unique ID Listing for a particular

message or all the messages.
ALL A macro that can be used in place of

<msgnum>. Assumes the verb LIST has been
previously run.
Sample Script
The following is sample script for running the POP3 module.
REM. Assumes that there are at least 10 items in the Inbox.
CONNECT
STLS
LOGON
STAT
LIST
UIDL RANDNUMBER(1,10)
RETR RANDNUMBER(1,9)
TOP 1 10
DELE 4
SLEEP RANDNUMBER(0,20)
RSET
NOOP
QUIT
The SMTP Module

The Simple Mail Transfer Protocol (SMTP) module simulates multiple clients using the SMTP
protocol. The module supports parameters for configuring the following:
69
 Options for how the test executes
Examples of options are the server that this client is targeting or the port number. Client actions
are simulated by running a script that contains scripted module keywords and SMTP verbs.
are logged.
Script Verbs Specific to the SMTP Module

The following table lists script verbs and their functions.
AUTH [LOGIN] Authenticates the current user by using the

prescribed method.
BDAT <file path> Sends the specified file.

specified port.
DATA <file path> Sends the specified data.
EHLO <string> Sends the EHLO command.
HELO <string> Sends the HELO command.
HELP Sends the Help command.
MAIL FROM <smtp address> Sends the MAIL FROM command. Defaults to
current user.
NOOP Do nothing.
RCPT TO <smtp address> Sends the RCPT TO command. Defaults to a

random user in the user group.
RSET Resets the session.
QUIT Ends the session.
STARTTLS Begins a TLS session.
VRFY <smtp address> Verifies the SMTP address.
Sample Script
The following is sample script for running the SMTP module.
CONNECT
70
EHLO ADMIN
STARTTLS
EHLO
AUTH NTLM
MAIL FROM: user.emailaddress
RCPT TO: SEQUUSER.emailaddress
RCPT TO: RANDUSER.emailaddress
RCPT TO: NEARBY.emailaddress
RCPT TO:
%40 RCPT TO:
%30 RCPT TO:
BDAT RANDLIST(d:\loadgen\perfbh\Exiser-Perf-BH-Inbound-messages.txt)
NOOP
QUIT
Sleep RANDNUMBER(500,1000)
The IMAP4 Module

The Internet Message Access Protocol version 4 (IMAP4) module simulates multiple clients using
the IMAP4 protocol. The module supports parameters for configuring the following:
 Options for how the test executes
Examples of options are the server that this client is targeting or the port number. Client actions
are simulated by running a script that contains scripted module keywords and IMAP4 verbs.
are logged.
Script Verbs Specific to the IMAP4 Module

The following table lists the script verbs and their functions.
APPEND <mailbox name> <file name> Appends the contents of <filename> to the
mailbox.
AUTHENTICATE <package name> Authenticates the connection as the current

user.
71
CAPABILITY Requests the capability string from the server.
CHECK Performs a Check command against the server

(equivalent to a NOOP in Exchange).
CLOSE Closes the selected folder. This deletes any

items that are flagged for deletion in the folder.

specified port.
COPY <num><dest> Copies an item or set of items from the

selected mailbox to a destination mailbox.
CREATE <mailbox name> Creates a specified mailbox.
DELETE <mailbox name> Deletes a specified mailbox.
DONE Executes the Done command.
EXAMINE <mailbox name> Examines a specified mailbox.
EXPUNGE <mailbox name> Calls the Expunge command on the selected

folder.
FETCH <num><msg data item string> Returns the item or set of items.
FETCHUID Translates to a “FETCH 1:* UID”.
IDLE Executes the Idle command.
LIST <reference name><wildcard> Lists folders.
LOGIN Authenticates the current user with the login

command.
LOGOUT Logs out the current user.
LSUB <reference name><wildcard> Lists subscribed folders.
NOOP Do nothing.
RENAME <mailbox name><new name> Renames the object.
SEARCH Executes the Search command.
SELECT <mailbox name> Specifies the selected mailbox.
STARTTLS Begins a TLS session.
STATUS <mailbox name> Returns the status of a specified mailbox and

its attributes.
STORE <num><msg data item name><item Stores the specified item.
72
value>
SUBSCRIBE <mailbox name> Stores the specified item.
UIDCOPY <num><dest mailbox> Copies the specified object to a destination.
UIDFETCH <num><msg data item set> Uses the UID to fetch an object.
UIDSTORE <num><msg data item set> Stores a specified object.
UNSUBSCRIBE <mailbox name> Unsubscribes to a specified mailbox.
Sample Script
The following is sample script for running the IMAP4 module.
REM # This script assumes the presence of at least 10 message currently
in the users INBOX
REM # And will create said structure
REM # This script will leave the INBOX with the first 10 messages of the
INBOX remaining... all others will be purged.
CONNECT
STARTTLS
REM LOGIN
AUTHENTICATE NTLM
CAPABILITY
SELECT "INBOX"
LOOP 10
APPEND "INBOX" .\scripts\file1.txt
ENDLOOP
LOGOUT
CONNECT
STARTTLS
LOGIN
SELECT "INBOX"
CREATE "BackMeUp"
COPY 1:10 "BackMeUp"
73
CREATE "Other Mail"
CREATE "Trash"
IDLE
DONE
LIST "" "*"
SELECT "INBOX"
FETCHUID
UIDFETCH UIDALL (UID RFC822.SIZE rfc822.header)
NOOP
UIDFETCH FIRST 1 (UID RFC822.SIZE RFC822)
UIDCOPY 1 "Trash"
UIDSTORE 1 +FLAGS (\Deleted)
UIDFETCH NEXT (UID RFC822.SIZE RFC822)
UIDCOPY 2 "Other Mail"
UIDCOPY 1 "Trash"
UIDCOPY 2 "Other Mail"
UIDCOPY 1 "Trash"
CHECK
NOOP
74
FETCH LAST 1 (UID)
UIDCOPY LAST 2 "Other Mail"
UIDSTORE LAST 2 +FLAGS (\Deleted)
CLOSE
SELECT "Other Mail"
FETCHUID
UIDFETCH 1:* (FLAGS)
NOOP
UIDFETCH FIRST 1 (UID RFC822.SIZE RFC822)
FETCHUID
UIDCOPY LAST 4 "Trash"
UIDSTORE LAST 4 +FLAGS (\Deleted)
CLOSE
SELECT "INBOX"
FETCHUID
UIDFETCH 1:10 (FLAGS)
NOOP
UIDCOPY LAST 5 "Trash"
UIDSTORE UIDALL +FLAGS (\Deleted)
EXPUNGE
CLOSE
SELECT "TRASH"
STORE ALL +FLAGS (\DELETED)
75
EXPUNGE
LIST "TRASH" "*"
DELETE "Other Mail"
DELETE "TRASH"
CLOSE
SELECT "BackMeUp"
COPY 1:10 "INBOX"
DELETE "BackMeUp"
LOGOUT
SKIP(12)
The Outlook Web Access Module

For the Outlook Web Access module, test cases are written in script files that specify the actions
to be done for each stress test user. The test module parses and reads this script file and
executes each command by building and sending the appropriate HTTP requests. Each
command consists of the name of the action to be performed and a series of parameters for the
action (if applicable).
This section describes the Outlook Web Access module commands.
AssignPolicyToItem Command
The AssignPolicyToItem command simulates assigning a policy to the active item by adjusting the
appropriate policy property tag on the item. A random simulated policy is assigned to the item as
a result of this command. The command syntax is as follows:
AssignPolicyToItem
Auth Command
The Auth command sends authentication tokens to the server as part of the logon process. The
command syntax is as follows:
Auth
CreateAppointment Command
The CreateAppointment command creates an appointment on the user’s calendar. The command
syntax is as follows:
CreateAppointment
CreateAttachmentFromForm Command
The CreateAttachmentFromForm command uploads a simulated attachment to the in scope item.
The command syntax is as follows:
76
CreateAttachmentFromForm
CreateFolder Command
The CreateFolder command creates a subfolder beneath a specified folder. The name of the
created folder is generated by the module and cannot be manipulated. The command syntax is
as follows:
CreateFolder <parent folder path>
An example of command usage is as follows:
CreateFolder Inbox
CreateSaveItem Command
The CreateSaveItem command creates a new item to be persisted in the mailbox. The command
CreateSaveItem
CreateSendItem Command
The CreateSendItem command creates a new item to be persisted in the mailbox and sent to
configured recipients. The command syntax is as follows:
CreateSendItem
DeleteConversation Command
The DeleteConversation command deletes items associated with a conversation. The default
behavior of the command with no arguments will delete the active conversation. Alternatively, the
additional parameter “random” will cause the command to delete a random conversation. The
DeleteConversation
DeleteConversation random
DeleteItem Command
The DeleteItem command deletes an item from the active conversation. The command syntax is
as follows:
DeleteItem
FindAllCachedPeople Command
The FindAllCachedPeople command queries all people in the user’s recipient cache. The
FindAllCachedPeople
77
FindCalendarItem Command
The FindCalendarItem command finds calendar items within a given view of the calendar relative
to the current date. Acceptable parameters include “day”, “workweek”, “week”, and “month”. The
FindCalendarItem <calendar view>
FindCalendarItem workweek
FindConversation Command
The FindConversation command finds a conversation. The command syntax is as follows:
FindConversation
FindFolder Command
The FindFolder command finds folders. The command syntax is as follows:
FindFolder
FindItem Command
The FindItem command finds items contained in the specified folder. The command syntax is as
follows:
FindItem <folder path>
FindItem Inbox
FindPeople Command
The FindPeople command finds people using the specified search string. The command syntax is
as follows:
FindPeople <search string>
FindPeople SomeString
FindSenderItself Command
The FindSenderItself command finds the user associated with the OWA session. The command
FindSenderItself
GetAuthenticationPage Command
78
The GetAuthenticationPage command requests the OWA logon page as part of the logon
sequence. The command syntax is as follows:
GetAuthenticationPage
GetCalendarFolders Command
The GetCalendarFolders command retrieves all calendar folders for a user. The command syntax
is as follows:
GetCalendarFolders
GetCalendarItem Command
The GetCalendarItem command retrieves details for a random calendar item in the mailbox. The
GetCalendarItem
GetConversationItems Command
The GetConversationItems command retrieves details on the items associated with a
conversation. The default behavior of the command with no arguments will use the active
conversation. Alternatively, the additional parameter “random” will cause the command to use a
random conversation. The command syntax is as follows:
GetConversationItems
GetConversationItems random
GetFavorites Command
The GetFavorites command retrives the favorite folder list. The command syntax is as follows:
GetFavorites
GetFolder Command
The GetFolder command simulates opening the folder specified by the folder name and sets this
folder as the active folder.
The command syntax is as follows:
OpenFolder <folder name>
OpenFolder Inbox
GetItem Command
The GetItem command retrieves details associated with the active item. The command syntax is
as follows:
79
GetItem
GetMailTips Command
The GetMailTips command retrieves MailTips for the currently active person. The command
GetMailTips
GetOwaUserConfiguration Command
The GetOwaUserConfiguration command retrieves OWA configuration information for the current
user. The command syntax is as follows:
GetOwaUserConfiguration
GetOwaUserOofSettings Command
The GetOwaUserOofSettings command retrieves OOF settings for the current user. The
GetOwaUserOofSettings
GetPersonaPhoto Command
The GetPersonaPhoto command retrieves a photo of the currently active person. The command
GetPersonaPhoto
GetReminders Command
The GetReminders command retrieves reminders. The command syntax is as follows:
GetReminders
GetSessionData Command
The GetSessionData command retrieves data from the server to initialize an OWA session. The
GetSessionData
GetUserAvailability Command
The GetUserAvailability command retrieves availability information for the current user in the
specified timeframe relative to the current time. Acceptable parameters include “day”,
“workweek”, “week”, and “month”. The command syntax is as follows:
GetUserAvailability <timeframe>
80
GetUserAvailability week
Logoff Command
The Logoff command ends the OWA session for the user. The command syntax is as follows:
Logoff
Logon Command
The Logon command completes the logon sequence and typically follows the
GetAuthenticationPage and Auth commands. The command syntax is as follows:
Logon
MarkConversationAsRead Command
The MarkConversationAsRead command marks the currently active conversation as read. The
MarkConversationAsRead
MoveConversation Command
The MoveConversation command moves the active conversation to a subfolder under the current
folder. The command syntax is as follows:
MoveConversation
Search Command
The Search command searches for a specified sring. The command syntax is as follows:
Search <search string>
Search SomeString
UpdateAppointment Command
The UpdateAppointment command updates the current appointment. The command syntax is as
follows:
UpdateAppointment
UpdateItem Command
The UpdateItem command updates the current item. The command syntax is as follows:
UpdateItem
UserSpecificResourceInjector Command
81
The UserSpecificResourceInjector command submits a request to retrieve appropriate resources
for the OWA session. This is part of the typical process of establishing a new OWA session. The
UserSpecificResourceInjector
82
Copyright
The information contained in this document represents the current view of Microsoft Corporation
on the issues discussed as of the date of publication. Because Microsoft must respond to
changing market conditions, it should not be interpreted to be a commitment on the part of
Microsoft, and Microsoft cannot guarantee the accuracy of any information presented after the
date of publication.
This White Paper is for informational purposes only. MICROSOFT MAKES NO WARRANTIES,
EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS DOCUMENT.
Complying with all applicable copyright laws is the responsibility of the user. Without limiting the
rights under copyright, no part of this document may be reproduced, stored in or introduced into a
retrieval system, or transmitted in any form or by any means (electronic, mechanical,
photocopying, recording, or otherwise), or for any purpose, without the express written permission
of Microsoft Corporation.
Microsoft may have patents, patent applications, trademarks, copyrights, or other intellectual
property rights covering subject matter in this document. Except as expressly provided in any
written license agreement from Microsoft, the furnishing of this document does not give you any
license to these patents, trademarks, copyrights, or other intellectual property.
Unless otherwise noted, the companies, organizations, products, domain names, e-mail
addresses, logos, people, places, and events depicted in examples herein are fictitious. No
association with any real company, organization, product, domain name, e-mail address, logo,
person, place, or event is intended or should be inferred.
© 2013 Microsoft Corporation. All rights reserved.
Microsoft, MS-DOS, Windows, Windows Server, Windows Vista, Active Directory, ActiveSync,
ActiveX, Entourage, Excel, FrontPage, Hotmail, JScript, Microsoft Press, MSDN, MSN, Outlook,
SharePoint, Visual Basic, Visual C++, Visual Studio, Win32, Windows Mobile, Windows NT, and
Windows Server System are either registered trademarks or trademarks of Microsoft Corporation
in the United States and/or other countries.
All other trademarks are property of their respective owners.
83

Microsoft Exchange Load Generator 2013

Uploaded by

Copyright:

Available Formats

Microsoft Exchange Load Generator 2013

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Microsoft Exchange Load Generator 2013

Uploaded by

Copyright:

Available Formats

Microsoft® Exchange Load Generator 2013

System Requirements for Exchange Load

Component Minimum Recommended

Processor Any of the following Any of the following

Operating system Microsoft Windows Server Microsoft Windows Server

Memory 512 megabytes (MB) of RAM 2 gigabytes (GB) of RAM

Available hard-disk drive 300 MB 300 MB

Display SVGA (800 x 600) or higher SVGA (800 x 600) or higher

Running against Exchange Exchange Server 2013 Exchange Server 2013

When to Use Exchange Load Generator

Installing Exchange Load Generator on a Client

To install Exchange Load Generator on the client computer

1. How to Configure Exchange RPC/HTTP. Proper configuration of RPC/HTTP connections

Steps and sub steps Method(s) used

Starting the simulation Start menu, graphical user interface, or

Setting up the topology, which involves Graphical user interface

Customizing the test by task Edit the configuration file parameters

Common task: Enabling stress mode Edit the configuration file

Common task: Customizing the initialization of Edit the configuration file

Enabling remote control Edit the configuration file. Graphical user

Setting up the test topology Graphical user interface

Initializing the test Edit the configuration file

Monitoring the progress of the initialization Run Performance Monitor (Perfmon)

Initializing the mailboxes Edit the configuration file

Initializing public folders Edit the configuration file

Running the test Command-line program. Graphical user

Stopping the simulation test Command-line program. Graphical user

Repeating the test Control Panel > Administrative Tools. Active

How to Configure Exchange RPC/HTTP

To configure an external Outlook Anywhere namespace

Get-OutlookAnywhere |Set-OutlookAnywhere -ExternalHostname mail.contoso.com -

To install a certificate on CAS servers

4. Save the file and close Notepad.

To validate that Outlook Anywhere connectivity is working

How to Start Exchange Load Generator

To start Exchange Load Generator on the Start menu

To start Exchange Load Generator at a command prompt

How to Set Up the Exchange Load Generator Test Topology

To set up the topology for an Exchange Load Generator test

To verify that users are created in Active Directory

To configure the test properties

Customizing the Exchange Load Generator Test

Tables of Simulation Tasks and Their Parameters

MinimumActionMessages The minimum number of calendar items to

MaximumActionMessages The maximum number of calendar items to

ChangePercentage The percentage of calendar items to be

DeletePercentage The percentage of calendar items to be

FolderOpenCount The number of folders that will be opened.

OldPostReadCount The number of old posts to read, in addition to

CreateSubFolderPercentage The percentage of task invocations in which a

DeleteSubFolderPercentage The percentage of task invocations in which a

MinimumFolderDepth The minimum folder depth for the public folder

MaximumFolderDepth The maximum folder depth for the public folder

AverageFolderDepth The average folder depth for the public folder

ResolveConflictsPercentage The percentage of conflicts that must be

ApplyViewPercentage The percentage of folder opens in which a

ReplyPercentage The percentage of processed messages that

ReplyToFolderPercentage The percentage of processed messages that

ForwardPercentage The percentage of processed messages that

DeletePercentage The percentage of processed messages that